Hacker News Reader

Comment by TallGuyShort | Hacker News Reader

Correct me if any of the following is wrong, though: when we say the universe is expanding, we mean that matter is spreading out. The further apart the matter is, the faster it's spreading out. If we back-trace the position of stars, they all coincide at roughly the same point in time and space. This led us to believe that at some point in time all of this matter was together, incredibly dense, and in a very "hot", high-energy state. This led to the prediction that has since been confirmed, that from the furthest reaches of the observable universe, we would see cosmic background radiation. And as time goes, that cosmic background radiation is getting further and further away, so the observable bubble is expanding.

Maybe I have some of that wrong, but that's my understanding. None of that means that the fabric of space (as in space-time) is actually expanding. And even if it did, none of that means that nothing could have existed outside of the observable bubble billions of years ago, and could have since been passed by the observable edge and now be inside that observable bubble. Unless maybe the source of the background radiation was still dense / high-energy enough to absolutely obliterate everything in it's path. And maybe it was? But I just don't see how we can reason about what was outside of the observable bubble.

View on HN

Perhaps a big bang that's 'just' an explosion in spacetime, rather than of spacetime, would not work with known physics because that concentration of that much matter would be a black hole? Until now, I never thought to ask anyone.

I'm not a physicist but I suspect this comment hits the nail on the head. You can only "undo" a black-hole by expanding the space it is embedded in.

I think string theory approaches this idea with the holographic principle: https://en.m.wikipedia.org/wiki/Holographic_principle

It wouldn't work out because it wouldn't match what our telescopes are observing. This image has a good example of what's happening

https://upload.wikimedia.org/wikipedia/commons/c/c7/Space_Ex...

You can see that from any given point:

1. Every other point is moving away from you

2. Points further away are moving away at a faster rate

This could only work out if space were expanding. If it were an explosion instead, everything would be moving away from a fixed central point.

That actually demonstrates the opposite: despite the fact that the images themselves show the expansion of a grid in a two-dimensional space, each of the dots appears, from its perspective, to be at the center of the expansion, unless you either specify an absolute frame of reference, or (which amounts to the same thing) can identify the edge of the expanding region.

Furthermore, our telescopes have only observed from the vicinity of the sun - we have not actually seen what the universe looks like from another star. Of course, we can figure out what it would look like, but if that is your argument, then it is circular.

> That actually demonstrates the opposite: despite the fact that the images themselves show the expansion of a grid in a two-dimensional space, each of the dots appears, from its perspective, to be at the center of the expansion, unless you either specify an absolute frame of reference, or (which amounts to the same thing) can identify the edge of the expanding region.

The image isn't enough to demonstrate the difference. But the speed of light is, along with the way acceleration and redshift work.

> Of course, we can figure out what it would look like, but if that is your argument, then it is circular.

A chain of logic that starts with the evidence we have, goes on to what telescopes show elsewhere, and then concludes about expansion, is not circular.

You might have some specific circular argument in mind, but there are non-circular arguments for what GP is saying.

> The image isn't enough to demonstrate the difference...

Precisely.

> ...But the speed of light is, along with the way acceleration and redshift work.

Are you referring to the cosmological component of redshift? I suppose that, at least in principle, we could determine the cosmological component by first measuring the Hubble constant from the change in distance, over time, of those stars whose distance can be measured without any assumptions about redshift, but has that actually been done? (or some other experiment that directly calculates the cosmological component of redshift?) OP's claim is that we already have sufficient experimental results, not that we could, in principle, get them. AFAIK, experimental error in these stellar distance measurements is too large, and they are over too short a period of time, for the rate of expansion of the universe to be taken directly from them (note that the anomaly that is the subject of the article itself brings into question the accuracy of the cosmic distance ladder.)

> A chain of logic that starts with the evidence we have, goes on to what telescopes show elsewhere, and then concludes about expansion, is not circular.

It seems pretty clear that OP was thinking that if we were to observe the universe from a distant galaxy, it also would seem to be at the center of the expansion - but that would be the case in both scenarios, so it does not actually distinguish between them. If, however, OP had in mind some other set of measurements that, when combined with those from Earth, distinguish between the scenarios, then the question becomes, can you deduce what those far-away measurements would be without choosing between the two scenarios, either explicitly or implicitly? If that choice is being made, then you are begging the question. What set of measurements and calculation do you have in mind that avoid this circularity? (the above-mentioned redshift measurements are a separate issue, both because they are clearly not what OP had in mind, and because they do not depend on any calculation of what observations would be made from a distant galaxy.)

> Precisely.

Demonstrating the opposite is an very different thing from failing to demonstrating a difference at all.

> What set of measurements and calculation do you have in mind that avoid this circularity?

Any measurements that would actually work, I think.

You're the one asserting circularity, so I think it's up to you to explain what specific argument would be circular. Especially when you're apparently talking about theoretical arguments that swebs might have.

At best, you are claiming that swebs' argument was correct because there is a different argument for the point he claimed has been demonstrated - but that is, as you say, "a very different thing." My original objection to that argument stands, and you have agreed that it does!

> Demonstrating the opposite is an very different thing from failing to demonstrating a difference at all.

Who is making any claim about demonstrating the opposite? You can see from my original post that I agree that spacetime is expanding; I merely disagree with swebs' argument for claiming to know that it is.

>> What set of measurements and calculation do you have in mind that avoid this circularity?

> Any measurements that would actually work, I think.

In that case, you will have no difficulty in stating, or providing citations for, the actual measurements and calculation that get the job done - if any such set of measurements exist, that is. Otherwise, your reply is equivalent to "I don't know."

My point is that I do not think that there is any set of deduced remote measurements that a) show, as a certainty, that space itself is expanding, and b) can be deduced without implicitly or explicitly assuming a position on the issue. Current cosmology expects that the large-scale measurements will show the same results as from Earth (e.g. the same Hubble constant, if and only if you derive it using the same assumption about space itself expanding), but the absence of a difference between there and Earth would fail to satisfy a).

Furthermore, I notice that you have still not made clear which redshift measurements support your claim that the issue has been resolved experimentally. A citation would be sufficient.

> Who is making any claim about demonstrating the opposite?

I guess I misunderstood when you said "That actually demonstrates the opposite"...

I probably could have been more explicit. To be clear, the opposites here are 'the big bang is an expansion of spacetime' and 'the big bang is an expansion in spacetime.' Swebs claimed the diagram showed the former, but it acually literally shows an expansion in space (which, it so happens, also works for an expansion of space, and so does not, as it stands, show a way to choose between the two.)

Let's go over the circularity issue in more detail. It starts with Swebs' statement that "[The big bang as a uniform expansion in spacetime] wouldn't work out because it wouldn't match what our telescopes are observing." Here we have a claim that the 'uniform expansion in spacetime' hypothesis (UEIS for short), which we all agree is wrong, has already been ruled out by astronomical observation.

Next, Swebs offers a specific argument for that claim, in the form of a diagram which shows that, in a uniformly expanding universe, each point appears to be that from which the expansion is ocurrring. One problem with that argument, which we both apparently agree on, is that this would be true for UEIS (for one thing, the image can be taken as a diagram of a demonstration using marbles on a table, which would literally be a UEIS.)

There is a second problem with it, however: none of our telescopes have observed the universe from the perspective of a distant galaxy, so we cannot say that our telescopes have shown that everywhere seems to be at the center of the expansion. We can deduce what it looks like from distant galaxies, if and only if we make some assumptions about the dynamics of the universe (there may be other observations from Earth that have already ruled out UEIS empirically, in which case the rest of this comment is moot. You have claimed that there are, but so far, you have have not presented any details.)

There are two possibilities: either it looks the same as from Earth (putting aside details local to the specific point of observation) or there is some difference. In the former case, which would hold if the expansion we observe is a uniform expansion of spacetime (UEOS), and which therefore is what we all, and cosmology in general, assumes to be so, I do not think these observations would rule out the possibility of UEIS. Therefore, for observations made in a distant galaxy to empirically settle the UEIS/UEOS issue, they must differ from those from Earth, in a way that is diagnostic of, and therefore causally dependent on, what form the expansion takes. Furthermore, because the local and remote results would differ in this case, UEOS would be ruled out.

So, when we try to deduce what can be observed from a distant galaxy, we must either assume that it is the same as from Earth (subject to local corrections), which will fail to resolve the issue, or we must deduce that there is some difference - but what difference? As shown in the previous paragraph, the relevant differences must be causally dependent on what form the expansion takes, so you have to make assumptions about the latter in order to deduce the former. There's the circularity, and it is general - i.e. not dependent on what specific measurements are being considered.

> You have claimed that there are, but so far, you have have not presented any details.

You can measure stars and watch the distances accelerate, or look at how the redshift of a star changes over time instead of staying static. These might take a while to do, but this is about what we can measure, not what we already have.

> As shown in the previous paragraph, the relevant differences must be causally dependent on what form the expansion takes, so you have to make assumptions about the latter in order to deduce the former.

This is that part where I see a breakdown in your logic. It's very possible that we can make measurements on Earth that would tell us exactly how our telescopes would differ from a distant galaxy. In that case, there would be no circularity.

Note that the proposition "We can only get this data by going to another galaxy and using telescopes there." is not part of this scenario.

> This is about what we can measure, not what we already have.

That is not so, this discussion started with a claim that we already have observational proof that distinguishes between the two scenarios. I have already suggested that the Hubble constant could, in principle, be measured without cosmological assumptions, but we do not, AFAIK, have good enough data, and you have not offered anything to alter that opinion.

>... look at how the redshift of a star changes over time instead of staying static.

Up until now, you have been very vague about what sort of measurements you have had in mind. This is getting closer, but under which scenario does the redshift stay static?

> This is that part where I see a breakdown in your logic...

All you are doing here is disagreeing with the conclusion; you have not offered any refutation of the argument that leads to it. The reason for me saying "the relevant differences must be causally dependent on what form the expansion takes" is that if they were not, then they would not be the sort of differences that could be used to choose between the alternatives. And if they were so dependent, then the mapping from what we observe from here to what would be observed there also has that dependency.

> It's very possible that we can make measurements on Earth that would tell us exactly how our telescopes would differ from a distant galaxy.

There is something very confused here, but I cannot figure out what it is. I can say that my point does not have anything to do with "how our telescopes would differ from a distant galaxy."

> Note that the proposition "We can only get this data by going to another galaxy and using telescopes there." is not part of this scenario.

Actually, Swebs' original claim was pretty much that, as it could accurately be paraphrased as "we have seen, with our telescopes, that, when observed from a different galaxy, the expansion also seems to have originated there." You seem to have put yourself into the position of defending a claim that you do not actually believe in.

Note that there are some conditionals here that you have to be careful of. What I said was that there are certain claims that, if they were made, would result in a circular argument. I did not say that Swebs had made those claims, as he had not done so; I wrote it in an attempt to forestall a trip down a dead end (so much for that!)

There's definitely some deep miscommunication going on. I'm just going to try to clarify a couple parts...

> All you are doing here is disagreeing with the conclusion; you have not offered any refutation of the argument that leads to it. The reason for me saying "the relevant differences must be causally dependent on what form the expansion takes" is that if they were not, then they would not be the sort of differences that could be used to choose between the alternatives. And if they were so dependent, then the mapping from what we observe from here to what would be observed there also has that dependency.

It's circular if we assume the form of the expansion. If the real form influences our measurements, and we use that to figure out what telescopes would show, which lets us calculate the real form, then nothing is circular.

> Actually, Swebs' original claim was pretty much that, as it could accurately be paraphrased as "we have seen, with our telescopes, that, when observed from a different galaxy, the expansion also seems to have originated there." You seem to have put yourself into the position of defending a claim that you do not actually believe in.

No, those are very different statements. "We can tell from here what telescopes would show" could be part of a valid argument. "We need to go there to know" is not necessary.

> There is something very confused here, but I cannot figure out what it is. I can say that my point does not have anything to do with "how our telescopes would differ from a distant galaxy."

I'm talking about the "none of our telescopes have observed the universe from the perspective of a distant galaxy" stuff.

Okay, look, this is clearly not working. If you want to format your argument as a numbered chain of logical statements, I can probably give you a response you'll understand. Otherwise I'm giving up. Big blobs of paragraphs are not conducive to debating whether there "must" be certain assumptions.

I had written a long response to your various points, but this seems to get to the crux of the matter:

> It's circular if we assume the form of the expansion...

Yes.

> ...If the real form influences our measurements, and we use that to figure out what telescopes would show...

In what way is that not exactly the sort of circularity that we just agreed about above, given that the real form is the issue to be decided?

> In what way is that not exactly the sort of circularity that we just agreed about above, given that the real form is the issue to be decided?

Because measurements are not assumptions.

If we use assumptions about UEOS/UEIS to figure out what telescopes show, and use that as evidence for UEOS/UEIS, that's a circular argument.

If we use measurements to figure out what telescopes show, without any assumptions about UEOS/UEIS, and then use that as evidence for UEOS/UEIS, that's not a circular argument.

(And obviously there are always some assumptions when being sufficiently pedantic. It's an assumption that the sun still exists, etc. That's why I'm specifically saying "assumptions about UEOS/UEIS".)

> If we use assumptions about UEOS/UEIS to figure out what telescopes show, and use that as evidence for UEOS/UEIS, that's a circular argument.

Exactly, and that is what my comment to Swebs was intended to forestall (I have already made this point, several posts back.) That comment, being a reply to Swebs, must be read in that context (obviously, as you had not even joined the thread yet), and in that post, Swebs was making an argument that UEOS/UEIS has already been settled by observation. In other words, the resolution of UEOS/UEIS was, in fact, the conclusion that Swebs was claiming to have observational proof of. Using a pre-existing resolution of UEOS/UEIS to deduce what observations might be made elsewhere simply didn't enter into the discussion, as in that case, the issue being debated would have already been resolved.

> Because measurements are not assumptions.

I don't have measurements, Swebs didn't have measurements (though he thought he did), and it has become clear that you don't have measurements either. Maybe someone does, but they have not yet shown up in this discussion.

> in that post, Swebs was making an argument that UEOS/UEIS has already been settled by observation.

At most that was just being wrong, not having a circular argument.

> In other words, the resolution of UEOS/UEIS was, in fact, the conclusion that Swebs was claiming to have observational proof of.

Yes, Swebs was making a conclusion about UEOS/UEIS. But both the circular and non-circular arguments do that. Swebs was not using assumptions about UEOS/UEIS to reach that conclusion.

> I don't have measurements, Swebs didn't have measurements (though he thought he did), and it has become clear that you don't have measurements either. Maybe someone does, but they have not yet shown up in this discussion.

That doesn't matter.

If I say "A implies B", and I have no evidence for A, then I have not proven B. But it is not a circular argument.

You can have an argument that is both valid and unproven.

The argument "Measurements will/might/do tell us UEOS/UEIS" is not circular.

>> in that post, Swebs was making an argument that UEOS/UEIS has already been settled by observation.

> At most that was just being wrong, not having a circular argument.

I did not call Swebs' argument circular. I will repeat, for at least the third time now, that my comment about circularity was intended to forestall the use of a circular argument, of the sort that you have acknowledged is possible, in attempts to correct it.

>> I don't have measurements, Swebs didn't have measurements (though he thought he did), and it has become clear that you don't have measurements either. Maybe someone does, but they have not yet shown up in this discussion.

> That doesn't matter...

It is not intended to be an argument for the circularity of anything, it is simply a reply to your comment about measurements, pointing out the lack of them in this discussion.

> I did not call Swebs' argument circular.

I know. Swebs has not proposed a full argument. We're talking about hypothetical arguments that swebs, or someone else, could propose. Right?

> I will repeat, for at least the third time now, that my comment about circularity was intended to forestall the use of a circular argument, of the sort that you have acknowledged is possible, in attempts to correct it.

Here are the things you said:

"Of course, we can figure out what it would look like, but if that is your argument, then it is circular."

"you have to make assumptions about the latter in order to deduce the former"

"I do not think that there is any set of deduced remote measurements that a) show, as a certainty, that space itself is expanding, and b) can be deduced without implicitly or explicitly assuming a position on the issue."

"What I said was that there are certain claims that, if they were made, would result in a circular argument. I did not say that Swebs had made those claims, as he had not done so; I wrote it in an attempt to forestall a trip down a dead end (so much for that!)"

So, correct me if this summary is wrong, "Any argument based on figuring out what it looks like has to be a circular argument."

Do you stand by that statement, or am I misinterpreting something?

Because that's what I disagree with. I think it's possible to make an argument, based on figuring out what it looks like, that is not circular. This argument would use measurements as the basis for its figuring/deducing.

> circular argument, of the sort that you have acknowledged is possible

It's always possible to make a circular argument. What matters is whether it's possible to make a non-circular argument.

I am assuming that, in your statement of my position, where you wrote "figuring out what it looks like" you meant "figuring out what it looks like elsewhere." I also want to make it clear that I take "based on" to mean that the deduced remote measurements are a necessary part of the argument. Nevertheless, it is still not quite there: as it stands, even after these adjustments, it leads to the conclusion that Swebs' argument is circular, yet we are both agreed that it is not.

This goes back to the part of Swebs' argument that says, in effect, "if we were to observe the expansion of the universe from a distant galaxy, it would look like it does from here" (specifically, that we are at the center in both cases.) My position is that I do not think one could make any successful argument of that form - i.e. which depends on comparing what one would measure at a remote location with what one observes here - without it being circular. My reason for thinking that is so is that I suspect that, if the argument is valid (Swebs' was not), you cannot deduce what those remote measurements are without begging the question.

Now we get to the critical part: my reason for thinking it would be begging the question is that, if the measurements are capable of resolving the UEIS/UEOS issue, then I think they must have a causal dependency on which way UEIS/UEOS goes, and thus, on account of that dependency, one could not deduce what would be measured remotely without first picking one or the other.

Update: Let's look at it from the other direction: if your deduced remote measurements do not differ depending on whether you assume UEIS or UEOS, then how are they going to resolve the issue? This is what makes Swebs' argument an invalid one.

Note that this does not rule out UEIS/UEOS being settled by local measurements, and in fact, early in this thread, I mentioned what I thought might be a possible candidate: directly determine the Hubble constant without reference to redshift, through measurements over time of objects whose distance can be measured by other means. Such methods would not depend on comparing local and deduced remote measurements, and would render moot the whole issue of whether we can deduce relevant remote measurements, as well as all arguments that rely on doing so.

So my position is that I do not think there are any arguments that are both valid and non-circular, that depend on comparing deduced remote measurements to local ones, that show we already have observations that resolve the issue, and I will stand by it until I am persuaded otherwise. One valid counter-example would be sufficient.

Okay, I understand everything you're saying, so let me put this as simply as possible.

Let's posit that we're taking local measurements and using them to determine UEIS/UEOS.

We know it's possible to do a whole bunch of abstract math and get an answer, without involving remote galaxies.

But maybe that's hard math to do.

What if it's easier to extrapolate our measurements to a remote galaxy?

So we calculate what telescopes would show in a remote galaxy as an intermediate step.

And from that we actually determine UEIS/UEOS.

That would be a non-circular argument, in which the deduced remote measurements are a necessary step.

It's not the only argument we could have made, but it's a useful and valid one.

Is that hypothetical argument detailed enough to satisfy you? It's something that should apply to just about any measurements. There's an unlimited number of ways to prove anything. There's no reason "deduced remote telescope measurements" can't be an intermediate step for all sorts of proofs that look at stars for data.

> Update: Let's look at it from the other direction: if your deduced remote measurements do not differ depending on whether you assume UEIS or UEOS, then how are they going to resolve the issue? This is what makes Swebs' argument an invalid one.

I know that my local measurements will differ based on UEIS/UEOS. But I don't know specifically how. Just looking at my local measurements by themselves, I have no idea which one they support. And in the process of deducing the remote measurements, I don't assume either one. I just plug in the numbers.

If you want a math analogy, then uhhh think of the local measurements as a composite number, and factoring it to deduce the remote measurements? And the question we ultimately want to answer can easily be solved with the prime factors.

It would be possible to calculate the answer without factoring, but much harder.

And factoring does not depend on assuming the answer. It's just factoring. It depends only on the composite number.

I am not sure that I have made every implication of my position clear to you, as the arguments in the paragraph beginning "now we get to the critical part", and even more so in the update following it, imply that if there is some calculation where one can just plug in the numbers without explicitly or implicitly taking a position on UEIS/UEOS, it will not yield a result that resolves the issue (unless it merely duplicates a set of measurements that could just as well have been made locally, in which case the excercise would have been pointless.)

I think the best thing we can do at this point is to agree to disagree, especially as we are disagreeing over a hypothetical set of measurements and a calculation that neither of us can specify.

> I think the best thing we can do at this point is to agree to disagree

I would say yes to this except for your parenthetical there. Reading that, I think we're finally on the same page.

> unless it merely duplicates a set of measurements that could just as well have been made locally, in which case the excercise would have been pointless

Yes, it does duplicate the measurements that could have been made locally. But to whoever is writing the proof it's easier to use them in the 'remote galaxy' form. So it's not pointless. It's a critical part of that particular proof.

So with that noted down, I think everything is resolved?

I'm fine with that.

> I do not see how differing telescopes have anything to do with the issue.

How our telescope's observations would differ if they were placed in a different galaxy.

> it would have to be done without any tacit assumption about whether UEIS or UEOS applies

Yes! Now we're getting somewhere.

When you say "the argument is circular" that depends on all possible arguments using assumptions about UEIS/UEOS. I'm saying that you can make arguments that don't make assumptions about UEIS/UEOS. It's possible to make a circular or non-circular argument.

> In what way is that not exactly the sort of circularity that we just agreed about above, given that the real form is the issue to be decided?

Because I'm talking about measurements! You know, the same thing telescopes do. They measure. If you make measurements instead of assuming, then you don't have a circular argument.

> Mainly because you haven't said what measurements you think will substitute for Swebs' "the expansion appears to originate from wherever we are."

I don't have any in mind. I'm not saying there are any. I'm just saying that if the argument exists, it can be expressed in a non-circular form.

> point out the first statement in there that you disagree with

"We can deduce what it looks like from distant galaxies, if and only if we make some assumptions about the dynamics of the universe"

If "some assumptions about the dynamics" is supposed to include UEIS/UEOS assumptions, then this is where I disagree. We do not necessarily have to make those assumptions to make those deductions. You have not given any proof that all possible ways to deduce what it looks like will require those assumptions. We may be able to remove all of those assumptions with the right local measurements. Then we would have a non-circular argument.

> when we say the universe is expanding, we mean that matter is spreading out

It's been a while, but I recall reading in a book that's actually the space between everything that's expanding, much like you state in your second paragraph. For example, if you blow up a balloon, two points on it will appear to be moving apart from each other even though they're stationary.

I believe this Forbes article is in agreement: https://www.forbes.com/sites/startswithabang/2018/02/10/ask-...

It's also been described like a loaf of bread expanding, where stars are like raisins, except the loaf might be infinitely big. See https://en.wikipedia.org/wiki/Inflation_(cosmology) and https://en.wikipedia.org/wiki/Shape_of_the_universe

But that's also somewhat misleading as it gives the impression that it's just interstellar space (reinforcing the non-scientific connotation of space) that inflates. At some point even atoms may not be able to hold together. See https://en.wikipedia.org/wiki/Big_Rip Whether the Big Rip is true or not, it's useful as a supplemental illustration.

To be clear, the big rip is only if expansion gets massively faster. At the current rate, the disturbance on matter is ridiculously negligibly small.

Space itself is actually growing as well as celestial bodies having some relative motion that on average moves away from a common point. The growth in space itself is called inflation. Last time I studied cosmology (university) this was an accepted fact although the precise mechanism that explains this inflation was not agreed upon (scalar field, dark energy etc). I think there is some debate about whether inflation is speeding up or slowing down or how long it has occurred for but AFAIK the concensus is that it is a real thing.

To make an analogy, it is as if you had an elastic sheet with objects moving on its surface. The elastic sheet can be expanded which causes the distance between any two fixed points to grow.

(it's been many years since I studied physics so any corrections to my comments are appreciated)

Reply on news.ycombinator.com