[johnspannenburg (OP)]

OK. Total newbie here but I think I have an 'issue' with the horizon problem as it is stated on pretty much every website that I look at. 

Here is a direct extract from wikipedia regarding the horizon problem:

"If one were to look at a galaxy ten billion light years away in one direction, say "west", and another in the opposite direction, "east", the total distance between them is twenty billion light years."


Before I get to that problem let me ask a question.

Assuming we could see through the Cosmic Microwave Background Radiation (CMB) basically right to the big bang (assume we can detect and view gravitational waves say and thus observe beyond the CMB) - The question is.. In which direction would I have to look to observe the universe (and in order to avoid some complication) when it was say the size of a marble or some other similar very small (non zero) size?

I believe the answer to that question leads inescapably to conclusions which undermine the whole basis for the 'horizon problem' but I would like some comments or answers to the question first.

Cheers

JS

[Soul Surfer]

Like the cosmic microwave background you will see back to the start of the universe in all directions even back to when it was the size of a marble but you will probably only see a small part of the marble because the rest is outside of your light cone

[johnspannenburg (OP)]

OK - I agree with that totally. 

Just like the CMB you will see the start in all directions.
This may seem counter-intuitive at first as you have (effectively) a point source (and compared to the size of the universe now the CMB is also pretty close to a point source) and yet it comes from every direction.
 
Personally I visualise this by mentally mapping 3d space onto the surface of a sphere that is expanding in time.  Then think of a point of light as it makes it's way across the surface (actually 3d space) - when it reaches it's destination the distance it has traveled is larger than the initial distance due to expansion of the universe. 

Anyway cutting that ramble short - what this surely means is that all directions from the observer (us on earth say) tend toward the exact same direction as you look further away and hence backward in time. 
At very small distances you can ignore these effects and if I walk 10km one way and someone else 10km in the opposite direction we will surely be 20km (approx) apart.  However when you start looking further in distance and time, as per the quote in my previous post from wikipedia, then surely that whole statement of the 'horizon problem' is simply a nonsense. We look 10 Billion light years in one direction and then we look 10 billion light years in the other direction and the two points we see are plainly not 20 billion light years apart - as these directions are not opposite!  We are looking in almost the same direction. 

So if we look (with our gravity scope) back 13.7B years in opposite directions we are effectively looking at the exact same point (the "marble") although the two views would be looking at (not exactly but close enough for this example) opposite sides of the 'marble'(more properly points on the marble opposite each other but within our light cone).  Does this not destroy the whole basis for the 'horizon problem'?

[Soul Surfer]

No that is precisely what is meant by space being "curved"

You must remember that the points outside the point that you are looking at on the marble are further away than light can travel.  the universe is not the size of a marble for very long and light probably only travelled a tiny fraction of the distance across the marble in the time betwee it was that size and twice that size.

Remember also that we are talking about a four dimensional space-time geometry and it is very difficult to think in four dimensions you are not looking a the surface of an object. switch back to thinking in three dimensions.  You are at some random point INSIDE the object with everything around you expanding.  the spots on an inflating balloon image has serious limitations.

[johnspannenburg (OP)]

I don't understand what you mean but it's clear that I was not clear enough also.
Let me rephrase.

In my expanding sphere example (where the surface of that sphere is actually 3d space and from the centre out through the surface is time) take a point on the marble (E) and say that's the point that will one day expand to become where earth is now. 

On the marble you can draw a circle which represents the point beyond which light can never reach E.  This is the horizon.  We will never see anything of the marble outside that circle.  That "circle", to us, of course is seen as a spherical shell around our position.

Now take two points  A and B on the marble which are just inside the horizon but which are on opposite sides of E.  They are just enough inside the horizon that they just happen to reach E now.  What we might now think we see are two photons coming from A and B which appear to be each 13.7B (say) light years apart. 

In fact they are NOT that far apart as the photons (when the light was emitted) were very very much closer together.  As we take this example further back to where the marble is (almost) a single point.. the two photons are ,effectively, coming from the same point.

Again we have what appears to be two photons coming from opposite sides (and in a simplistic view they have) but those two 'opposite' points were so close together in the past they , now, they are effectively coming from so close to the same direction as to not really make any difference.

If you think of the light cone which represents our horizon (the cone is what the expanding circle on the marble traces out over time) then what we actually see of the universe can be plotted as a shape within that cone staring at the centre of the expanding base (which is us)and then curving asymptotically towards the walls of the cone as we work (backwards in time) towards the apex.

Again the inevitable conclusion is that the further back we look the more we are looking at a smaller and smaller universe (and ultimately a single point) and that two photons appearing come from opposite directions each 13.7B LY away is simply an incorrect assumption based on a 3d perspective of what is really a 4d universe.

Hoping this clarifies somewhat.
Cheers.
JS

[Soul Surfer]

You seem to have got the idea pretty exactly but the thing that you see as a problem is precisely the way it is and OK.  This is what the "inflationary universe"is all about the bits of our universe that we see as the cosmic microwave background were so close together originally were virtually the same place and subjected to the same conditions.

Now a strict mathematical singularity as the origin does not make physical sense (to me and many scientists  and this model will break down at some time and this breakdown will probably reveal very significant new ideas of our universe.  The model will be a merger of gravitation and standard quantum theory and maybe even a quantum theory of gravity possibly taking aspects of string theory into the model.  I have posted some suggestions as to how I personally think this might work in the new theories topic under "evolutionary cosmology".

http://www.thenakedscientists.com/forum/index.php?topic=11668.0

But I stress that there are other ideas but most of them offer no way of proving them scientifically.  The big advantage of what I am suggesting is that it might just be accessible to modelling and experimentation so the idea could be disprovable like the old continuous creation cosmology.  Disproving it could result in a real advance in science as did the disproving of the CC theory.

[johnspannenburg (OP)]

I'm reading a couple of papers regarding some the fundamental topics in this discussion but from what I see so far my problem with the way the 'horizon problem' is stated remains.
 
At the very least when distances are stated it should be specified what distances are being used 'proper' or 'comoving' (or other as appropriate) as without this being specified the discussion really becomes meaningless - as you no doubt are aware.  I've discovered that distance measures in cosmology is a topic all in itself (and quite interesting). I guess what it all boils down to is that the way the 'horizon problem' is stated almost everywhere is, at best, incomplete.

Cheers.
JS

[tychobrahe]

If those two points come from one almost singular point 13.7 billion years ago, so do we, and we'd just be the point in between those two almost touching points.  Those points are also radiating light in the other direction but all that light is out of our light cone and unknowable.  When the universe was infinitely dense and undifferentiated, it was still infinite.  I think you may be confusing the limits of the observable universe with the limits of the universe.  The marble you are imagining is the observable universe at that time, not the entire infinite universe, and those two points would be relatively as distant as they are now as they were then, at the very edges of the light cones only then it was a much smaller light cone.  There is no proof for or against the shape of the universe being positively curved, so it's probably not safe to assume it is with a mental image projected on a sphere.

[johnspannenburg (OP)]

Well no I'm definitely not confusing those two things. 

The 'marble' I'm imagining is the entire universe and the circle drawn upon it is the limit of the observable universe - which would actually appear as a sphere to us - in fact the 'hubble sphere' as I now understand the terminology.

The entire marble = the entire universe at that time and my concept of the "infinite" universe (in my little mental analogy and in absence of proof which would destroy it) is that if you were able to travel around the surface of that marble then you would of course never get to any 'end' you may get back to where you started if you fully circumnavigate but there is no end.  Hence this universe is finite but unbounded - no end but not infinite.

Of course you can not actually do this as, thinking about the horizon or hubble sphere in reverse, it also represents the maximum point to which you could ever hope to travel. It is the point where the recessional speed = c and therefore (unless you can travel faster then light) it would take infinite time to get there.  In practice then - the universe does become , in a manner, infinite because most of it is too far away for us to ever reach. 

Anything on the hubble sphere would take infinity to reach and anything beyond that... well still infinity I guess!  ;-)

Cheers.
JS

[johnspannenburg (OP)]

So, Soul Surfer, are you able to clarify the 'horizon problem'?  Taking the wikipedia 'definition' as an example what do you believe the "10 billion light years" distance really is?  Is it the distance between where the photon started the journey and where E was at the same cosmological time?

That is, is it the proper distance from A(past) to E(past)  or is it the proper distance from A(now) (which we can't see of course) to E(now) or is it the comoving distance which , as I understand it, remains constant and can be factored up by the "scale factor" of the universe at a given time to arrive at proper distance? (somewhat simplified definition).

Thanks.

JS

[johnspannenburg (OP)]

Can anyone shed some light / more properly state or give an example of the 'horizon problem'?

Cheers.
JS

[Soul Surfer]

I have reread the conversation trying to identify where your conceptual problem lies and I think I now understand it.  

It is vitally important to realise that the expansion of the universe from the big bang is the expansion of space itself and not the relative velocities of the galaxies after a big explosion.  The expansion of space is the one thing that allows things to move apart faster than the velocity of light and although those two points were much closer in the distant past they were always separating from each other faster than the velocity of light.  This is definitely not the same as two objects coming out of a big explosion each moving apart through space with the three quarters of the velocity of light which can see each other moving apart at over 90% of the velocity of light but definitely not one and a half times the velocity of light.  The actual relative velocities of all the galaxies through space in the universe are typically a few hundreds of miles per second towards or away from each other and nowhere near the velocity of light.

[johnspannenburg (OP)]

Thanks for the reply but these are not the source of my problem.
I believe the velocity between galaxies you are referring to is known at the 'peculiar' velocity and for my example two objects with a peculiar separation velocity of zero will work just fine.
The two objects or points I was referring to were (between themselves) certainly moving apart at greater than the speed of light but not seperating from E (us) at greater than the speed of light as the points were just inside the horizon and the light got to us for us to be able to observe it.

Anyway the summary of my problem with the 'horizon' problem remains with the fact that when it is stated (everywhere) that I could find it simply talks about 10 billion light years (or whatever distance they choose) when the term 'light year' without specifing 'proper' or 'comoving' distance only makes sense on a very local scale. 

The distance needs to be properly specified when you are working on cosmological distance and time scales.  When we see the photons arrive from some very far away point (time and space) it does not make sense to simply say 10B light years. 

In the time the photon took to get here space has expanded tremendously and that photon emitter is no longer where it was.  So the 10B refers to what? the distance between where the emitter was at time of emission of the photon and where we were at the same cosmological time?  Or is it the distance between us and where the emitter is now? at this cosmological time?  Of course the emitter may not even exist 'now' anymore but we will not know that until sufficient time as passed for photons (if any) from the emmiter 'now' reach us at some point in the future.

In summary then.. ignore peculiar velocities altogether.
When they say "10 billion light years away" what does that distance mean..?
Is it 'proper' now, 'proper' then, 'comoving' distance?

Hopefully this clears up the core of my query.. ;-)

Cheers.
JS

[Soul Surfer]

A light year is always the distance light travels during a year so it is essentially a measure of time and then movement at the velocity of light.   The objects have obviously moved in the time between the light being emitted and being received.  I see no problem with this.  This measure is used because it is the easiest way of describing things and because it represents things as we see them which is the only thing that we can be reasonably sure of.  Everything else must be some form of extrapolation.

The distance of galaxies is estimated mainly by red shift and this brings the Hubble constant into it.  There are possible errors and inconstancies in this but on the whole the picture seems quite good Type 1a supernovae give an alternative yardstick and there are other more abstruse measuring techniques being developed.

[johnspannenburg (OP)]

Yes I understand and pretty much agree with all of that but...
It does nothing to address my question which is in relation the the statement of the horizon problem.

For the problem to make any sense the distance must be properly stated.  If, when stating the distance as 'x' lightyears they really mean the 'comoving' distance or some other distance then this has to be clarified for the statement to have any meaning.

I've read several papers including

"Expanding Confusion:  common misconceptions of cosmological horizons and the superluminal expansion of the universe" by Tamara M. Davis and Charles H. Lineweaver

and

Distance measures in cosmology by David W. Hogg.

Both of these are very interesting read and will perhaps clarify what my 'issue' is.  It is simply not sufficient to say a 'lightyear' when you are referring to highly redshifted objects.  Perhaps there is an implied term or descriptor (as to distance) when the horizon problem is stated - if so maybe someone who understands this could explain it.

You mention redshift but even that is not quite a simple as just 'redshift'.  Redshift really consists of two components.  These two components derive from (1)the 'peculiar' velocity of the object in question and (2)the expansion of the universe (known as the cosmological redshift).  There are many situations where knowing which you are referring to or understanding the breakdown of the two components is crucial to the issue.

Similarly, I believe, the horizon problem (as stated in wikipedia for example) is either incorrectly stated or not stated in sufficiently technical terms for it (to me) to make proper sense.

I hope I get to the bottom of this... LOL.

Cheers.
JS

[Soul Surfer]

Redshift is only consistently detectable when it is larger than the basic motions of the galaxies but in the case of galaxy clusters which would have similar redshift the basic radial velocities (what you for some reason call peculiar) can be analysed out statistically.

I disagree with you about your approach to distance and time measurement you appear to be wanting to understand what the universe looks like at one particular moment and not the way we and everyone else sees it through a "time tube" created by the velocity of light.  Your problems are therefore ones of your own creation and not of the generally accepted approach.

[johnspannenburg (OP)]

Peculiar is not my term it's a term used to describe relative motions other than caused by the expansion of the universe itself.  Just as 'comoving' is a technical term used in cosmology

If you refer to the papers I mentioned you will understand.

Reading those papers might also shed some further light on 'redshift'
None of these ideas are my approach - they are simply what I am reading in scientific papers published in recognised places.

I do not want to understand, per se,  what the universe looks like at one particular moment.  We all, unavoidably see it through a time tube as you call it.. Although many do NOT understand this.

All I wish is a fuller explanation of the horizon problem using more scientific terms than what are typically used - nothing more and nothing less.  Again reference to the papers I mentioned might further your understanding, at the least of what exactly I'm trying to get to the bottom of.

Please don't misunderstand my comments - I'm sure you are very knowledgeable in this area but it would appear that the particular question I'm trying to get to the bottom of requires  more understanding of the precise cosmological terminology than  perhaps either of us have.

We're not really making any progress.  If you were to read the two papers I mentioned we may have ground for further discussion.

Cheers.

[LeeE]

I've got to admit that I just don't 'get' the Horizon Problem, at least as presented in the wikipedia article, even though it seems as good an explanation of it as any.

...One would expect, then, that their physical properties would be different, and more generally, that the universe as a whole would have varying properties in different areas.

What I don't understand is why it is assumed that things should be different when there's no reason for them to be so.  If different parts of the universe were formed under the same initial conditions, as implied by the BB, then there would need to be an additional factor to bring about a subsequent difference between the separated parts after they had separated.  Moreover, this additional factor must have also had its origins in the BB, so why would it only apply to one part and not the other?

In any case, I can't see any references to such an additional factor that would only apply to one part and not the other.

Problem?  What problem? []

[johnspannenburg (OP)]

OK.  As I understand it.

There is no reason things should be different - in fact the issue is that areas of space so far apart that they could NOT have had causal contact are SO similar that they MUST have had causal contact!!! 

These two statements (or the two parts of that one statement) seem to contradict each other but it comes down to the words 'causal contact'.  Normally this (loosely) means areas of space where a photon (or anything travelling at the speed of light) CAN travel from the one area and eventually get to the other area.  Or... if point B is within the Hubble Sphere of Point A then they are in causal contact and information etc can travel between the two and temperatures can equalise and so on...

Now the 'Horizon Problem' says (sort of - and I have my problem here) that there are parts of the universe so far apart that they can NOT have had causal contact (the speed of light thing) and yet they are so similar in temperature that they must have been..

(let's ignore your contention that they all came from the BB so what's the problem - for now) 

Cosmology solves the contradiction above by simply saying that there must have been a time when the universe expanded at some phenomenal rate (superluminal expansion) and that's how two areas of space so far apart can be so similar in temperature.. This is the theory of inflation.

In respect of your BB comment... my understanding is that quantum fluctuations way back then - even after the expansion of the universe should result in (for areas that can [and again my problem surfaces here] not have had causal contact) larger variations in temperature than we actually see.  So.... they must have had contact and we need inflation.. yada yada.

So.. my issue (in the way the horizon problem seems to be stated) is , in some ways similar to yours, that when wikipedia (say) says words to the effect of - look east and see some object 10B LY away and look opposite to West and see some other object 10 LY away and so those object are 20B LY apart and the universe is only 13.7B LY old and therefore we need inflation.

I say.. hang on... what is the 10B LY distance they are talking about.. 'proper' distance, 'comoving' distance or what? (please refer to my earlier posts regarding these terms)  without this clarification the Horizon Problem seems, at best, to be improperly stated.  I do not believe that East and West are, in effect 'opposite' when you take the effect of time and the expansion of the universe into account.

Taking my example to the extreme.. if you could , somehow, detect particles from right back to the BB (and they would be coming from EVERY direction EXACTLY the same as the CMB) then you are detecting particles, effectively, from the same point in spacetime.... despite the fact that the appear to come from 'opposite' directions.  In effect my contention is that when we look out, from our position in spacetime, in EVERY direction there is.. then wherever you look... that point.. ultimately would end up being back at the BB and would be the SAME point.  You've got to get your head around it but it means particles / photons apparently at 'opposite' end of the universe are NOT.. they are (more and more as you go back in time) actually in the same place! And so these 'supposedly' far apart sources are NOT far apart .. and why the need for inflation?!  They aren't causally seperated ... they are in the same place...(ultimately)..

Crikey I'm rambling.. hope that made some sort of sense..

Cheers.

[Soul Surfer]

I now see what part of the argument you are not happy about.

The horizon problem is a bit more complicated than that.  The real question is whether, when we look back into the past through the "time tube" we are seeing more of the universe less of the universe or exactly the same amount of the universe as time passes.  That is is space really as flat as it appears to be at first sight.  Quite clearly we cannot wait long enough to see the difference.

If it is flat then the points are aways separate right back to the singularity but as you say if it is a true singularity they are all exactly the same place.  If you see more the universe could be smaller than we can see and we could in some directions (maybe several) be looking at the back of our head round the curvature of the universe. People have been analysing the CMB for the correlated patterns that would be visible if this was true.  Some have suggested that they might have seen hints of such a pattern.  alternatively we may be seeing less meaning that the universe is totally open.

My personal opinion is that like the epicycles of the attempt to fit the Ptolemaic cosmos to observations we are lacking an important fact that will solve this problem.  To my mind mathematical singularities make no physical sense and that the expansion from a point breaks down before you reach a mathematical singularity.  I expect that most other workers in the field will agree with this.

Many experts argue that to do this we need a fully worked out theory of quantum gravity and string theory goes some way towards this with great mathematical complexity and a vast array of possibilities that are not scientifically testable.  My particular argument is that we can go some way to understanding how and where this breakdown might be without a fully worked out theory of quantum gravity by analysing in detail the processes involved in the collapse of a Kerr (rotating) black hole inside its event horizon.  Kerr said what the final state would be  a ring singularity but that's just the same as saying our universe ends with the heat death and says nothing about what happens on the way ie life etc.

Going beyond that we enter into the "new theories" area  and yo can find out more by looking at

http://www.thenakedscientists.com/forum/index.php?topic=11668.0