One topic which generated a lot of discussion at the Gravity and Cosmology meeting was the void model of the Universe. The basic argument is simple: the dark energy is an ugly addition to our cosmological standard model, with 70% of the energy density of the Universe some mysterious substance with weird properties. From a theoretical perspective, dark energy has the wrong density by many, many orders of magnitude, and worse, we may never be able to study it directly in the laboratory. Now suppose I told you I had a model which explained all of the observations, was based on general relativity, and appealed to no mysterious dark energy component (but still has dark matter, unfortunately). Sounds tempting, right? This is precisely what John Moffat, Chris Clarkson, Antonio Enea Romano, Chul-Moon Yoo, and others were advocating at the workshop (Kenji Tomita has also done a lot of work on this; the model has been around for decades). There’s one important caveat, however. The void model throws out the homogeneity and isotropy assumption. The Universe is now spherically symmetric, with a big hole in the middle. Even worse, we happen to be very, very close to the center of the hole.
As I discussed in a previous post, John Moffat argues that we shouldn’t be any more disturbed by this model than the standard model, because they’re both anti-Copernican: the void model in space, the standard model in time. As I discuss in that post, I’m not sure I completely agree with this. The fine tuning for the average void model is fairly involved. First, the matter density must be carefully set, as a function of radius, to agree with observation of the luminosity-distance relation. Then we have to be set down within roughly 1 Mpc of the center of the spherical void (which is at least a few Gpc on a side). If we were at a random spot in the Universe, there’s a probability of much less than 1 in 10 billion that we’d end up sufficiently close to the center of a void (assuming such voids existed). On the other hand, the standard Lambda CDM model of cosmology requires fine-tuning of the cosmological constant to a tiny, but non-zero number. To some this is unbearably ugly. But, at the end of the day, it’s just one additional, arbitrary number.
All this being said, what’s great about void models is that they aren’t just a philosophical alternative to the standard model. This is physics. There are measurements that can be done to differentiate (and possibly falsify) these models. Stebbins & Caldwell have come up with one particularly interesting approach, exploiting the fact that “random” observers in a void model see a different sky (and hence, a different CMB) from the one we do in our privileged position. It is surprising that a model so radically different from our standard model is still viable (although under pressure). Tests over the next few years are expected to distinguish these models, and we’ll know definitively whether we are at the center of the Universe.
July 29th, 2010 at 10:40 am
Oh holy Oreos. This theory is definitely going to be picked up by someone who understands it about as well as I do (which is to say, doesn’t understand it at all) and made into a truism about how we are
a. the center of the universe and so
b. We are the most important aspect of the universe and therefore
c. There is intelligent design and the whole universe was created to support life on earth.
Yup. its coming.
July 29th, 2010 at 11:30 am
How many galaxies are close to this void just like ours?
I rest my case.
July 29th, 2010 at 11:35 am
I wouldn’t particularly mind being a center of the universe, or the most important aspect of the universe, or an element in a universe intelligently designed to support life on earth, as long as those concepts extend from reliable evidence and not the wiggery of some self-described prophet or group. My life–and likely most people’s–would be far better with a real purpose and understanding of existence.
If it turns out that you really can show me an intelligent designer of spina bifida babies and paranoid schizophrenics, I can take it; I could really use a vengeful god from up the sky to will and mete me the tears that I shed. I could bear it, clench myself, and die, at least as well as some ‘ol pessimistic poet, as long as there was sufficient evidence to support the concept.
However, that’s a long way from a proposed alternative to the standard model. Let the nuts do what they want to with the hollowed universe idea (they will anyway)–that nuts distort evidence is not sufficient reason to abandon searching for it.
Also, please note that there are some of us for which oreos really are holy. Our tastes may be plebian but that is no reason to gratuitously slam our belief structure.
July 29th, 2010 at 11:42 am
Maybe there’s an explanation for the local void. We might be like the unhygienic baseball player who covered most of the outfield because none of his friends would tell him. The ancient cosmology depended on a similar, if inverse notion: the Earth was the center of the Universe, but not because it was a good place. It was, instead, something of a septic tank that naturally gathered all of reality’s dreck.
July 29th, 2010 at 12:58 pm
[...] This post was mentioned on Twitter by John Lobert, J.S.. J.S. said: Are you the center of the Universe? http://ow.ly/18lCba [...]
July 29th, 2010 at 2:18 pm
If both models can explain observations both models are viable candidates. Homogeneity and isotropy while appealing are only assumptions and cannot be strictly true anyway.
That said, neither model is particularly appealing so we should keep searching for better alternatives.
July 29th, 2010 at 2:20 pm
Why exactly is fine tuning a problem? If observations are better described by a special rather than generic location (or distribution), so be it.
I do not quite understand why physicists seem to favor homogeneous and isotropic models. In this case at least, simplicity is not the reason, so is it some kind of philosophical preference? It resembles Plato’s prescription that all celestial motions be reduced to uniform circular ones, which resulted in Ptolemy’s epicycles (dark energy?).
And I am not sure I understand “probability of much less than 1 in 10 billion that we’d end up sufficiently close to the center of a void”. Isn’t probability the same for whatever location we end up at, special or otherwise?
July 29th, 2010 at 2:42 pm
Sergiy,
The preference isn’t just philosophical. By definition, generic locations are much more common than special ones. It is then much more probable (before evidence is collected) that we are somewhere generic. If we find that we are not, we should try to find a reason why it is so (that is, a reason that our location shouldn’t just be drawn from a uniform probability distribution on the universe), although unlikely results will occur from time to time statistically, too. You compare this mode of thinking to Ptolemy’s, but circles are, in fact, a very special case of the conic section family of curves. The idea that all motions must be composed of these non-generic paths is actually much closer to the idea that we sit near the very centre of a giant void.
Re: probability, you’re right that for any individual location the probability is identical (to a first approximation), or at least that’s the assumption. The point is that “generic” locations (i.e., those not at the centre of an enormous void) are not distinguishable, while non-generic ones are. Since there are vastly more generic locations than non-generic ones, the probability of ending up in one of the typical places is correspondingly greater than the probability of ending up somewhere special.
July 29th, 2010 at 3:57 pm
Suppose someone is flipping coins and a sequence HHHHHHHHHHHH comes up. If she follows ‘generic first’ maxim then this is a puzzle requiring an explanation. A pseudo-problem is created. On the other hand, equally probable HTTHHHTTTHHT requires no explanation. Why not vice versa? The center of the giant void is only special to us, just like the first sequence, the universe itself does not distinguish.
I like the point about circles, but conic sections themselves are no generic curves. Kepler could have kept his original ovals, but ellipses, special though they are, fit Tycho’s data better. If the giant void fits (without ad hoc assumptions like dark energy) why shouldn’t it be adopted?
I am not thrilled about the giant void either. But dark energy smacks too much of Lorentz’s ether: a ghost with artificial properties invented to save an extra-observational preference.
July 29th, 2010 at 4:31 pm
The real reason for us to be at the center of a void in the center of the universe is: the universe is trying to flee in every direction from Chuck Norris.
That said, I’m really glad to be seeing alternative ‘structural’ models like this (and a few others I’ve seen over the last year) get some attention. Every model that we can make detailed enough to rule out is a step in the right direction. I’ll have to read the Stebbins & Caldwell paper, because from just the abstract I don’t understand how the CMB measurement could tell us anything.
July 29th, 2010 at 4:39 pm
The paper below was posted recently, and seems to have some analysis relevant to the topic:
http://arxiv.org/abs/1007.3725
There’s a corresponding discussion at Cosmo Coffee. It seems like these models are being tested now, and are found lacking.
July 29th, 2010 at 4:59 pm
Sergiy,
It’s true that unlikely things can and do happen in nature, including things much less likely than flipping HHHHHHHHHHHH. I noted that improbable results occur from time to time statistically in my original reply. Past a certain point, though, the more unlikely a particular observation is to have occurred randomly, the more we should look for a deeper explanation.
With your coin-flipping example, we can take a viewpoint like that in statistical mechanics. For 12-flip data sets (microstates) like the ones in question, define some parameter (macroscopic observable) D = number(heads)-number(tails). There is only one possible data set with D = 12, but there are 12!/(6!6!)=924 possible data sets with D = 0, as in your second example. The probability of observing D = 12 is 1/2^12, while the probability of observing D = 0 is almost 25%. The main idea I’m trying to get across is that HTTHHHTTTHHT “looks” generic in a very quantifiable way, despite individually being just as improbable as HHHHHHHHHHHH, and that being in the centre of a giant void looks special (and hence improbable) in the same way that HHHHHHHHHHHH does.
July 29th, 2010 at 5:31 pm
Sergiy:
If you’re betting against someone, and they flip heads 16 times in a row, are you going to check the coin for being weighted before the 17th flip, or are you going to keep on betting?
July 29th, 2010 at 6:01 pm
What about David L. Wiltshire’s Timescape model:
http://arxiv.org/pdf/0909.0749v2
It appears to be different from the void models as it derives from the known inhomogeneities in the LSS and is a “fractal bubble” model. It does not use a LTB spacetime like the void models critically examined in the Moss, Zibin and Scott reference of #11, rather it uses the growth of the inhomogeneities and a back reation.
The fractal bubble assumption seems to me not to violate the Copernican Principle for an observer in a local void of the LSS but then I am not a physicist.
July 29th, 2010 at 6:18 pm
Here’s an interesting site about a group of prospective children of Howard Hughes, trying to get DNA to compare with theirs.
http://howardhugheschildern.blogspot.com/
Here’s som interesting debate on the subject of intrique in the Howard Hughes matter.
http://www.howardhughesblog.com/
July 29th, 2010 at 6:26 pm
Eric:
Let’s define a different observable: it is 1 if the sequence is anti-palyndromic and 0 otherwise. Then it is HTTHHHTTTHHT which is special and HHHHHHHHHHHH which is not. You can always define an observable that makes what you choose special. The reasons we choose some observables over others are of course entirely anthropocentric, and any sense in which HHHHHHHHHHHH is special is also entirely anthropocentric. I guess, my point is that arguments against supposedly anthropocentric universes are themselves based on anthropocentrism. From universe’s ‘point of view’ center of the void is no better or worse than any other place.
bittergradstudent:
If I am betting against ‘someone’, of course I will check the coin. Because ‘someone’ is likely to have the same idea of generic/special as I do, and on top of that has vested interest in fabricating a particular kind of special. I am afraid, it is the underlying assumption of ‘someone behind the universe generator’ that creates a problem out of fine tuning. What looks like fine tuning to us is nothing special in any sense detached from humans.
July 29th, 2010 at 6:34 pm
“Are You the Center of the Universe?”
Why yes! Yes, in fact I am!
July 30th, 2010 at 12:47 am
“I do not quite understand why physicists seem to favor homogeneous and isotropic models. In this case at least, simplicity is not the reason, so is it some kind of philosophical preference?”
Originally, it was just a working assumption to simplify the maths. However, we now observe isotropy to a large degree (galaxies, radio sources, CMB), so this is no longer an assumption. In addition, observed isotropy implies homogeneity unless we are in a privileged position. So, either you believe in homogeneity or you believe we are in a privileged position.
July 30th, 2010 at 1:55 am
Q: How many geocentricists does it take to screw in a lightbulb?
A: One, they hold up the bulb, and the universe revolves around them.
July 30th, 2010 at 3:48 am
That doesn’t explain why the CC is then zero either, so why bother getting rid of it?
July 30th, 2010 at 4:59 am
Bee: “That doesn’t explain why the CC is then zero either, so why bother getting rid of it?”
The value of zero eliminates the parameter from the model meaning that there is no longer a need for an explanation. At least not any more then there is always a need to explain why one particular model works and others don’t.
A nonzero parameter describes a real process which operates in nature, explaining such processes is much more important then explaining hypothetical processes which don’t exist. The first group is (almost certainly) finite and of practical importance, the second group is infinite and of only philosophical importance.
Besides from research perspective it’s important to explore all models able to explain the data and then to differentiate between them based on further experiments. Settling on the first model which seems acceptable is a wrong approach.
July 30th, 2010 at 6:12 am
@Sergiy
If any identifiable order to the universe is just mankind imposing order on the universe and asking a pointless ‘why’, then why do science at all? I think that everyone is open to the possibility that the universe is ordered and that we’re at a special vantage place, but thinks that things would require less explaining if we were just in a generic spot in a homogenous universe.
But ultimately this is an answer for observation anyway.
July 30th, 2010 at 10:56 am
Hey Dan,
Moffat’s model reminded me of an idea of Jim Bjorken : he spent a couple of hours trying to explain it me (as a naive graduate student) in Sean’s office one afternoon.
He asked the question “What is the smallest spacetime we can imagine that would fit all our current cosmological observations?” The answer is that we are living in a super special spot in a minimal size spacetime. For simplicity, let’s imagine we observe that we are accelerating, so we are heading into dS space. Then, according to his “principle of minimal spacetime”, our worldline trace the center of the causal diamond of dS, but outside this causal diamond can basically be anything.
Of course, the principle predicts, observationally, nothing (since it is constructed by fiat to fit all observations). But I think, in the light of all this landscape business, maybe it can has some theoretical implications.
July 30th, 2010 at 11:50 am
Naturally we are at the center of the universe. Any point in an infinite universe can be seen as the center of the universe. There was no universal big bang. The universe is infinite in time and domain.
July 30th, 2010 at 12:10 pm
Quite a neat idea and coincidentally I was just musing about the same topic on my blog yesterday ( http://theastronomist.fieldofscience.com/2010/07/hubble-bubble.html ) and a friend sent me an email to show me this post. Uncanny similarity! I think the colloquial term for the void model should be Hubble Bubble.
Of course the Hubble bubble or void idea violates not only sensible cosmological ideals, but also recent precision data. The question for me that remains open is which model is more uncomfortably finely tuned – being at the center of a void or LCDM? Further, it is of note that fine tuning a single parameter in one model and several parameters in another model is not a fair comparison of the single-parameter to the multi-parameter models; how do you compare the probability of cosmological (cosmic variance limited?!) outcomes?
July 30th, 2010 at 3:12 pm
Andries, I assume your infinite time model has an explanation of Hydrogen/Helium abundances? And the cosmic microwave background?
July 30th, 2010 at 3:24 pm
What on earth do you mean by a random spot in the universe? Don’t you have to agree on a coordinate system and a probability distribution before you can make these kinds of statements? This seems like a problem with fine-tuning arguments generally and it would be nice if someone could explain to me how cosmologists deal with it. As an example of the kind of concern I am raising look up the Box-Muller transformation.
July 30th, 2010 at 4:08 pm
Anthropic principle + Fermi paradox: we exist to observe a local void because any other civilizations that might otherwise threaten us were wiped out eons ago in a massive intergalactic war.
That’s two birds with one stone, people. Occam’s razor says I’m right.
(edit: yes I know this is a flagrant violation of Occam’s razor. Sorry.)
July 30th, 2010 at 7:44 pm
As a matter of interest, has anyone looked at cluster formation/evolution in a void model??
I would have though that the dynamics of clustering as a function of redshift (or other statistics that respond to nonlinearity) might well work out differently in a void model (even if we are allowed to assume we are in the center) relative to a “standard” dark energy scenario.
July 30th, 2010 at 7:57 pm
Nothing unfortunate about it. Dark Matter is awesome!
July 30th, 2010 at 8:08 pm
Newtonian gravitation can be explained as the result of matter emitting some mysterious particles (gravitons maybe) that causes other matter to be attracted to it. The 1/r^2 law follows. With that picture, you can think of Einstein’s gravitation as a correction to Newton’s theory of the gravity particles. To get the right correction for Einstein’s gravity you have to assume that those gravitons interact with each other to make more gravitons. The Big Bang corresponds to the fact that as time goes on, matter is effected by gravitons from more and more distant matter. And Dark Energy corresponds to an increase in the effect of the distant matter due to its self interaction.
July 31st, 2010 at 3:36 am
Bizarre, not one mention of the Suyeav-Zeldovich effect.
If there was a ginormous void, there would be an enormous warm spot in the CMB.
This is not observed. I wonder what further special pleads are invoked to get around this.
July 31st, 2010 at 10:43 am
A giant void wouldn’t give a warm spot on the CMB if you were right in the center of the void. It would just be a monopole effect.
July 31st, 2010 at 11:11 am
I would like to remind cosmologists of something Mandelbrot pointed out decades ago.
Unbounded discrete fractal models are consistent with:
(1) Approximate statistical homogeneity on an infinite number of hierarchical scales,
(2) Intrinsic inhomogeneity as the more accurate global description of the hierarchy, and
(3) Nobody at the “center” of anything global, since all “centers” are local. By definition there is no overall “center” to an infinite self-similar system.
Such a system has maximum relativity. It is without absolute time, space or scale. There are no uniquely preferred times or directions or sizes, if one rigorously includes the entire infinite hierarchy in the relevant assessments.
July 31st, 2010 at 5:31 pm
Late night reading invariably loses some comprehension.
I don’t get the notion that the standard cosmological model violates homogeneity & isotropy in time. My understanding is that the violations are _local_, which is an allowed-for violation.
I also seem to recall that the analysis of CMB data essentially throws out monopole and dipole effects, as the calibration is to the CMB rest frame (dipole = 0). I’m not sure if there’s a way to fish out any monopole effect from the data. But I’m relatively inexperienced with this, as I only started digging deeply into how the CMB is analyzed in the last few weeks.
August 2nd, 2010 at 5:03 am
bittergradstudent, My infinite time and space model for the universe assumes that the macro macro appearance of the universe does not change over time. The hydrogen helium abundances have always been there, as well as the proportion of dark matter energy (relative to total energy) and the dominance of ordinary matter relative to antimatter. A further assumption is that this constancy is maintained by (presently unknown) energy cycles in the universe. I do not wish to imply that the presently accepted powerful and impressive standard models can be ignored, only that these models are not perfect and may need adaptation.
August 2nd, 2010 at 10:14 am
[...] more, see Are you the center of the Universe?, Hubble [...]
August 20th, 2010 at 7:26 pm
There is no center too the Universe. Dont believe Me .Just ask Sally Baloonist
September 8th, 2010 at 9:53 am
[...] A new cosmological theory eliminates the need for dark energy…at the expense of throwing out isotropy and homogeneity. [...]