PDA

View Full Version : "We can't see beyond the event horizon, so we have to trust all the abstract math"



Nereid
2008-Oct-15, 04:11 PM
From this (ATM) thread (http://www.bautforum.com/against-mainstream/79996-fun-gr-fans.html), specifically a post (http://www.bautforum.com/1343500-post41.html) by 37.1101 (extract):


And how do you suppose/suggest/etc anyone could determine "what is going on in black holes" (presumably, 'inside' the event horizon)?

Specifically, what sorts of experiments or observations do you think might be possible, in principle, that would lead to such "a better understanding"?Well, isn't that convenient. We can't see beyond the event horizon (Theoretical assumption) so we will just have to trust all the abstract math that gives us the famous 'point' with zero volume and infinite mass!

There was an excellent post by Ken G, and several good ones by others, in a recent Q&A thread that has disappeared, on the nature of modern physics, the relationship with 'reality', and the role of "all the abstract math".

This post, by 37.1101, provides what might be an excellent example to illustrate this.

With General Relativity (GR), we have a theory which has extraordinary explanatory and predictive power (see Clifford Will's latest review, for example). The theory predicts 'black holes', each of which has an 'event horizon', 'inside' which no external observer can 'see'*. One consequence of this is a complete^ inability to test any predictions concerning the physics 'inside' the event horizon of a black hole ... unless and until an extension or replacement for GR is developed which does allow such 'observation'.

One corollary of this is that anything goes! There are essentially no restrictions on whatever speculative physics you wish to write to describe what's inside the event horizon**.

But that's just my understanding; what do other BAUTians think, re GR, event horizons, black holes, etc?

* I'm using quote marks because the terms have particular, somewhat narrow and technical, meanings; if anyone would like any of them spelled out, please ask.

^ this is not entirely accurate, and the exceptions may be interesting to discuss ...

** other than mass, charge, and angular momentum ('spin')

stutefish
2008-Oct-15, 04:18 PM
*I'm using quote marks because the terms have particular, somewhat narrow and technical, meanings; if anyone would like any of them spelled out, please ask.
Pick me! Pick me!

I'm intensely curious to know the technical meanings of 'black holes', 'event horizon', 'inside', and 'see' in the parent post.


^ this is not entirely accurate, and the exceptions may be interesting to discuss ...
I'd also love to know more about these exceptions!

Nereid
2008-Oct-15, 04:37 PM
'black hole', 'event horizon', and 'inside' are defined within GR itself (for the purposes of my post), so any standard GR textbook would be the best place to dive deeper; would you like some rough, non-technical definitions (i.e. without the math)?

'see' narrowly means 'detect any emissions from inside the event horizon, whether light or non-zero mass particles'; more generally it means something like 'determine or infer anything about the physical properties of anything inside'.

There are three 'bulk' properties of a black hole that can be determined from outside: mass, spin, and electric charge, as I already noted. There's also the property of entropy, or information content, and the extent to which the Standard Model (of particle physics) can be added to GR to produce consistent predictions beyond GR itself. An example of the latter is Hawking radiation; 'black hole entropy' is worthy of a separate thread. There's probably more, but those are the main ones I think.

John Mendenhall
2008-Oct-15, 04:45 PM
One corollary of this is that anything goes! There are essentially no restrictions on whatever speculative physics you wish to write to describe what's inside the event horizon**.



Isn't the breakdown in our description at the singularity, rather than the event horizon? And the singularity is singular because we don't (yet) have the knowledge to describe it, rather than being something unknowable?

Hey, missed you for a while, Nereid. Keep posting.

Regards, John M.

Nereid
2008-Oct-15, 05:07 PM
Isn't the breakdown in our description at the singularity, rather than the event horizon? And the singularity is singular because we don't (yet) have the knowledge to describe it, rather than being something unknowable?

Hey, missed you for a while, Nereid. Keep posting.

Regards, John M.
That gets to (the heart of?) the point Ken G was making (well, one of them) ... certainly GR makes predictions about the physics 'inside' the event horizon of a black hole, except at the singularity; however, the point I am making is somewhat different: even if GR 'says' (predicts) something about the insides of BHs, there's no way to test what it says, because you can't see inside (with the caveats noted), even in principle.

And if you can't see (or, more generally, test), what does it matter what the theory says?

An analogy might be 'what are the laws of physics for the *unobservable* parts of the universe?'

tdvance
2008-Oct-15, 05:42 PM
I'd like to address "we can't see beyond the event horizon" and suggest that it's at least theoretically plausible that we could, at least fuzzily.

Hawking, in conceding a bet, suggested information for what goes into a black hole is not destroyed, but comes out in an encoded form in the Hawking radiation. The "encoding" procedure is essentially determined by what is "inside the event horizon"--up to quantum uncertainty, this is believed to be deterministic (and even with quantum uncertainty, is deterministic, just that we can't measure it accurately).

So, let's take the "sufficiently-advanced civilization" used often in thought experiments. These beings can capture or create a small black hole and exercise a certain amount of control over what goes into it. They also have super-supercomputers for predicting, to a certain degree of accuracy, how the Hawking radiation should change when a given test particle is dropped into the black hole at a given velocity from a given location. The beings can then measure the Hawking radiation and see how it compares with the computerized model, perhaps adjusting the model and the underlying theory accordingly.

In this way, one can, at least in principle, confirm or deny General Relativity or alternative theories within the event horizon.

Nereid
2008-Oct-15, 05:59 PM
I'd like to address "we can't see beyond the event horizon" and suggest that it's at least theoretically plausible that we could, at least fuzzily.

Hawking, in conceding a bet, suggested information for what goes into a black hole is not destroyed, but comes out in an encoded form in the Hawking radiation. The "encoding" procedure is essentially determined by what is "inside the event horizon"--up to quantum uncertainty, this is believed to be deterministic (and even with quantum uncertainty, is deterministic, just that we can't measure it accurately).

So, let's take the "sufficiently-advanced civilization" used often in thought experiments. These beings can capture or create a small black hole and exercise a certain amount of control over what goes into it. They also have super-supercomputers for predicting, to a certain degree of accuracy, how the Hawking radiation should change when a given test particle is dropped into the black hole at a given velocity from a given location. The beings can then measure the Hawking radiation and see how it compares with the computerized model, perhaps adjusting the model and the underlying theory accordingly.

In this way, one can, at least in principle, confirm or deny General Relativity or alternative theories within the event horizon.
Thanks very much! :)

As I understand it, Hawking radiation requires more physics than GR, and to the extent that it does, what any "sufficiently-advanced civilization" may measure, and conclude from their measurements, requires at least an implicit acceptance of at least some physics other than GR ...

Of course, the unambiguous observation of Hawking radiation would, it seems, strongly imply the ability to do some kinds of tests of GR inside the event horizon ... so roll on the day when such observations are reported! :D

Tim Thompson
2008-Oct-15, 06:24 PM
Trusting all that abstract math actually works surprisingly well. The universe behaves physically in surprising concordance with the principles of mathematics (see, for instance, The Mathematical Universe (http://adsabs.harvard.edu/abs/2008FoPh...38..101T) by Max Tegmark). Heisenberg's thoroughly surprising uncertainty principle in physics, which we know from experience to be physically valid, can be derived from pure mathematics, with no interference from physical reasoning or experience (i.e., Fourier Transforms and Uncertainty Relations (http://www.mathpages.com/home/kmath488/kmath488.htm)). So one should not lightly brush away "abstract mathematics" as if it were just an excuse rather than a reason for taking an idea seriously in physics.

In the specific case of black holes, the argument is over the top. We do rely on the mathematics of general relativity to tell us that there should be things we call "black holes" in the theoretical universe. But we rely on observation to tell us whether or not there actually are such things in the physically real universe. Black holes are unique in that they are the only compact objects which do not have a hard surface. If you fall into a black hole, you fall through the event horizon and vanish. If you fall into anything else that we know of or theorize, you smack into a hard surface and explode in cosmic glory when all that kinetic energy is released in the collision. So, matter falling into a black hole should react differently than matter falling into any other object, and that should make black holes observationally distinguishable from any other object. I think we have reached the point where we can now confidently claim to have observed the unique effect of the black hole event horizon and therefore observationally verify the existence of theoretically predicted black holes; i.e., Narayan & McClintock, 2008 (http://adsabs.harvard.edu/abs/2008NewAR..51..733N); Remillard, et al., 2006 (http://adsabs.harvard.edu/abs/2006ApJ...646..407R); Done & Gierlinski, 2003 (http://adsabs.harvard.edu/abs/2003MNRAS.342.1041D). Meanwhile, Doleman, et al., 2008 (http://adsabs.harvard.edu/abs/2008Natur.455...78D) show that millimeter wave interferometry is closing in fast on the spatial scale of the event horizon of the supermassive black hole at the center of the Milky Way. And Huang, et al., 2007 (http://adsabs.harvard.edu/abs/2007MNRAS.379..833H) show that we will be able to detect the predicted shadow effect around our own supermassive black hole. So it won't be long before we are able to confidently detect the event horizon of the great big black hole at the center of the Galaxy, as we already have detected the event horizon for smaller black holes, and show observationally that it is in fact a great big black hole, and not a great big something else.

Nereid
2008-Oct-15, 06:56 PM
Thanks for the good summary of the current state of observations wrt black holes Tim.

The thing which I understand 37 was primarily interested in was what happens inside the event horizon, at least in the quote I used in the OP*, and specifically what the 'reality' of the singularity in the middle of a BH is, based on 'the abstract math'; tdvance has opened up the door a tad on this, wrt Hawking radiation and black hole entropy.

There's a new (to this thread) aspect that your post introduces: if it ain't broke, don't fix it, or Occam's razor: since GR has passed so many experimental and observational tests so well, a conservative approach is to assume it works just as well inside the event horizon ... unless there's a good reason otherwise.

And of course in the case of the singularity at the heart of a BH, there are many 'good reasons otherwise'!

* there's no doubt he is highly sceptical of other aspects of GR wrt black holes, per his posts in the ATM thread, but I want to leave them aside, in this thread.

stutefish
2008-Oct-15, 10:44 PM
'black hole', 'event horizon', and 'inside' are defined within GR itself (for the purposes of my post), so any standard GR textbook would be the best place to dive deeper; would you like some rough, non-technical definitions (i.e. without the math)?
Yes, please :)

DrRocket
2008-Oct-16, 03:37 AM
I'd like to address "we can't see beyond the event horizon" and suggest that it's at least theoretically plausible that we could, at least fuzzily.

Hawking, in conceding a bet, suggested information for what goes into a black hole is not destroyed, but comes out in an encoded form in the Hawking radiation. The "encoding" procedure is essentially determined by what is "inside the event horizon"--up to quantum uncertainty, this is believed to be deterministic (and even with quantum uncertainty, is deterministic, just that we can't measure it accurately).

So, let's take the "sufficiently-advanced civilization" used often in thought experiments. These beings can capture or create a small black hole and exercise a certain amount of control over what goes into it. They also have super-supercomputers for predicting, to a certain degree of accuracy, how the Hawking radiation should change when a given test particle is dropped into the black hole at a given velocity from a given location. The beings can then measure the Hawking radiation and see how it compares with the computerized model, perhaps adjusting the model and the underlying theory accordingly.

In this way, one can, at least in principle, confirm or deny General Relativity or alternative theories within the event horizon.

Although Hawkind did concede the bet regarding information, it is my understanding that the specialists are not yet convinced. I believe that Thorne is withholding judgment.

I am still personally a bit confused as to how one "proves theorems" as does Hawking concerning quantum effects with regard to black holes when we lack a consistent theory combining general relativity and quantum mechanics.

Ken G
2008-Oct-16, 07:40 AM
I'll add a couple points to the mix. First of all, I don't think the Hawking bet speaks to the issue of what the conditions are below the event horizon, because the information coming out was what went in from "our side", and the idea (what little I grasp of it) is that the information is preserved-- that means that what happens to that information doesn't probe what is underneath the event horizon expressly because the information doesn't get changed there.

Moreover, I think the physics of what happens "inside an event horizon" is so difficult to determine, it is really a problem in metaphysics, or even the philosophy of science. Physics depends so strongly on testable predictions, it runs into a very interesting conundrum whenever testable predictions are fundamentally limited. Event horizons would seem to satisfy that criterion.

It is perhaps instructive to note that an event horizon might be kind of like doing an image-charge calculation to get the electric field outside a conductor. The image charge is an imaginary construct that meets the same boundary conditions on the surface of the conductor, so it gets the field right outside the conductor. The reason for the analogy is that if you were not allowed to do measurements inside the conductor, you could never know if there was really a charge there or not. So how do we do physics on a region we can only make inferences about, based on what we see elsewhere?

The whole business reminds me of the "Heisenberg cut" in his interpretation of quantum mechanics. The region behind the event horizon is a bit like the quantum domain, where we can only make observations from the other side of the "cut" and draw inferences about the quantum side. The rationalist approach is to build a quantum ontology based on the most natural extension of the mathematics that seems to work. Unfortunately, doing that results in some very bizarre conclusions, like having "many worlds" spring out of the woodwork to maintain the unitarity of the closed system that includes the observing apparatus. But we cannot observe the interaction of the apparatus and the quantum system, only the result of that interaction projected onto the apparatus. The rest we have to piece together using the mathematics, or we can apply an ad hoc bridging postulate and not worry about it (the Copenhagen approach). Why can we not simply do the same with an event horizon? i think it is purely a question of philosophy, the age-old empiricism versus rationalism.

showboat
2008-Oct-16, 08:05 AM
Inorganic or organic universe question.

For a example one could look at clouds in the sky and see them not freeze exactly but repeat the same position many times a second.

The next question is are clouds staying place or is the observer changing to match such a perception.

A valid a supposition as 'So, let's take the "sufficiently-advanced civilization" used often in thought experiments. These beings can capture or create a small black hole and exercise a certain amount of control over what goes into it. They also have super-supercomputers for predicting, to a certain degree of accuracy, how the Hawking radiation should change when a given test particle is dropped into the black hole at a given velocity from a given location. The beings can then measure the Hawking radiation and see how it compares with the computerized model, perhaps adjusting the model and the underlying theory accordingly.'

but perceptional like vision which is more convincing than math which is estoteric to many people.

So from a mechanical physic structure one moves to a biophysic structure of the universe.

Because a cognizant mind could see such.

DrRocket
2008-Oct-17, 03:45 AM
...

Moreover, I think the physics of what happens "inside an event horizon" is so difficult to determine, it is really a problem in metaphysics, or even the philosophy of science. Physics depends so strongly on testable predictions, it runs into a very interesting conundrum whenever testable predictions are fundamentally limited. Event horizons would seem to satisfy that criterion.

It is perhaps instructive to note that an event horizon might be kind of like doing an image-charge calculation to get the electric field outside a conductor. The image charge is an imaginary construct that meets the same boundary conditions on the surface of the conductor, so it gets the field right outside the conductor. The reason for the analogy is that if you were not allowed to do measurements inside the conductor, you could never know if there was really a charge there or not. So how do we do physics on a region we can only make inferences about, based on what we see elsewhere?

The whole business reminds me of the "Heisenberg cut" in his interpretation of quantum mechanics. The region behind the event horizon is a bit like the quantum domain, where we can only make observations from the other side of the "cut" and draw inferences about the quantum side. The rationalist approach is to build a quantum ontology based on the most natural extension of the mathematics that seems to work. Unfortunately, doing that results in some very bizarre conclusions, like having "many worlds" spring out of the woodwork to maintain the unitarity of the closed system that includes the observing apparatus. But we cannot observe the interaction of the apparatus and the quantum system, only the result of that interaction projected onto the apparatus. The rest we have to piece together using the mathematics, or we can apply an ad hoc bridging postulate and not worry about it (the Copenhagen approach). Why can we not simply do the same with an event horizon? i think it is purely a question of philosophy, the age-old empiricism versus rationalism.

I am afraid that I have to engage in [gasp !] philosophy to address this question. But I will try to provide what I at least think is a conclusion of a sort anyway.

It is not possible to directly observe what happens inside the event horizon of a black hole. At least it is not possible if general relativity is correct, and the existence of black holes is basically a result in a belief in the accuracy of general relativity in the first place -- although we are perhaps getting close to observational data that they exist.

But if you accept the validity of general relativity, and one might well accept it on the basis of those many observations that have been made that confirm the predictive capability of GR, then you can make sensible statements about what happens on the other side of the event horizon. GR permits some solutions that are sensible all the way to (but not at) the singularity. What you cannot do is obtain direct evidence that those solutions are valid.

But we use basic principles of physics, and general relativity is such a principle, rather often to make predictions that we do not directly measure. The difference is that we do not do so because it is inconvenient to do so or because the phenomena are difficult to isolate and measure directly. Instead we have performed controlled laboratory experiments that verify the accuracy of the principles under very carefully controlled conditions and we have made many other observations to validate the basic principles and the mathematical models based on them. That large body of experiment, observation, and theory provides the confidence to use the principles and models in many situations with complete confidence in the accuracy of the predictions. When those models break down it is a great surprise and cause to re-examine the validity of the principles themselves.

With respect to what happens on the other side of an event horizon we are confronted with two viewpoints. 1. It is in principle impossible to directly observe what happens on the other side (unless general relativity breaks down in ways that would be most surprising). 2. If general relativity is correct and if an event horizon has the properties predicted by that theory then it doesn't much matter what happens on the other side (one might quibble as to whether Hawking radiation happens on the other side or precisely at the boundary, but I am ignoring that at the moment, perhaps out of ignorance). But basically if I can't see it and it can't affect me the only reason to care about it is curiosity, and the only way I can address my curiosity is through application of general relativity.

I don't' think this is empiricism vs rationalism. I am forced to use what appears to be the rationalist approach, because I am unable, in principle, to perform the experiment that my empiricist side desires. On the other hand the theory that I am relying upon has my confidence only because of a huge body of empirical data that supports the theory as a whole. There is no conflict of desires here. I want to employ the mathematical reasoning of a rationalistic approach and I want to confirm it with the experiment of an empiricist. But I cannot. So, I take the pragmatic position of a [gasp !] engineer or politician and I do what I can do.

This is nothing new. I think that the notion that matter is fundamentally composed of atoms is pretty well accepted. There is pretty good evidence that it is true. Quantum mechanics and chemistry work pretty well. But the evidence is basically the fact that models and predictions based on the atomic model work exceedingly well. To the best of my knowledge no one has produced an image of an atom and its constituents (I have seen the images of crystals that show atomic positions, but they don't resolve the atom itself), and I have never seen an electron. But quantum theory tells me how atoms are constructed and, more or less, what goes on inside them. Similarly generally relativity tells me that black holes exist and describes space-time inside the event horizon (as long as I stay away from the singularity).

So, you can either believe your physical models, with the appropriate skepticism of a scientist who knows that all of our physical laws are but good approximations, or you can take the position that only that that I have actually seen and measured is real. But if you do the latter you set aside a great deal, I would think most, of modern physics. That might be OK for a philosopher, but it is rather career-limiting for a scientist. For the event horizon question I suppose that there is one other viable position -- Who cares, since it doesn't affect me? But that rather rejects curiosity which ought to be an anathema to either a philosopher or a physicist.

Ken G
2008-Oct-17, 04:27 AM
I am afraid that I have to engage in [gasp !] philosophy to address this question.The beauty of engaging in philosophy is that essentially everything is debatable. So although I see validity in what you are saying, I also see the counterpoint, and in some cases I actually favor the contrary view, for reasons I will be characteristically happy to expound on.


But if you accept the validity of general relativity, and one might well accept it on the basis of those many observations that have been made that confirm the predictive capability of GR, then you can make sensible statements about what happens on the other side of the event horizon. We can agree that GR is our best theory here, and there is no real point in considering specific competing theories for the purpose of this discussion, but even with the context of pure GR there is a problem in defining "accept the validity". A point I tend to stress is the importance of not reversing the logic of science: we must make our theories fit our observations, but we have only one reason to ask our observations to fit our theories-- that the theories are only standing in for other observations that have already been done. Note this latter requirement does not actually require any theory at all, it is just a shortcut-- we could as easily just mention the specific observations in detail, and thereby assert our contention that new "similar" observations have no reason to come out any different. That is the sole way in which we can rightly use theory to dictate to reality-- in all other ways, we must let reality dictate to theory or we are just not doing science.

So how does that principle apply to event horizons? The question is then, have we done observations that were interpreted as supporting GR that allow us to claim that what happens behind an event horizon is "similar" to what we have already observed, and used to build GR? That is the challenge that must be met, it is insufficient to simply say "GR explains all the observations we have been able to do, so everything else that GR says must be equally reliable."

The pitfall of using theory to dictate to reality is that you don't always know when you passed a limit of some kind, such that the new observations might not be "similar" to the set of old ones that verified the theory. A classic example of this is Newtonian mechanics-- what do you think Newton would have said if some young student had asked him, "but how do you know your theory will work to arbitrarily high speeds?" I suspect he might have said "as I have not the least inclination to believe anything special happens at any particular speed, I feel safe in that extrapolation". And note that in that situation, there wasn't any warning signs at all that the speed of light could be a critical boundary of sorts-- here we have an event horizon that our theory, GR, tells us is a boundary! That's a flashing warning sign-- Newton had far less reason to anticipate a velocity boundary for his theory, yet we do know that his theory runs afoul of a boundary all the same.


I don't' think this is empiricism vs rationalism. I am forced to use what appears to be the rationalist approach, because I am unable, in principle, to perform the experiment that my empiricist side desires. On the other hand the theory that I am relying upon has my confidence only because of a huge body of empirical data that supports the theory as a whole. There is no conflict of desires here.Tension between mpiricism vs. rationalism can be encountered in two ways, and I believe you are only focusing on one of them. Certainly, we can have situations where our instruments are telling us something different than our minds are ready to accept (like double-slit diffraction patterns in single-particle trajectories, or time dilation in a moving clock). The ancients might have thought there was a real concern there, but the modern view has come down pretty firmly on the side of empiricism-- when rationalistic thoughts don't agree with empirical evidence, reframe the thoughts, don't doubt the experiments. So that form of tension between empiricism and rationalism has already been put pretty much to rest, at least at the moment.

But there remains a far more insidious tension between rationalism and empiricism that appears not when the two seem to contradict each other, but rather when one is trying to decide the very meaning of empirical data. If I do an experiment in one situation, pure empiricism would tell me I have learned nothing about any other situation, because I have no way of knowing if some other situation might not entail some new variable I have overlooked that could change everything. Thus, pure empiricism is completely blind. Instead, we always stir in a dose of rationalism, at the very minimum to afford a useful general meaning to anything empirical.

But how much rationalism is needed to make empiricism mean something, and are we allowed to go any past that minimum requirement? That's what I view as the "real" tension between empiricism and rationalism, the fact that they both need each other to be of any use, but too much of the other would subjugate the first. There is a need for balance-- but what is the right balance? That is Nereid's question here, as I see it, in regard to an event horizon. It is an interesting angle from which to address that fundamental issue.

More in a moment.

publius
2008-Oct-17, 04:51 AM
THe Rindler horizon is also a boundary an accelerating observer can't see beyond. You're on a space-ship accelerating and you have been since forever. There's a big black horizon behind. A "infinite mass sheet" could lurk behind there. Or nothing could. :lol:

Is this the same question? And it's a matter of coordinates. But what's different about black holes in that regard?

There is the point that a local space-time, an observable space-time, can't tell you the global space-time. That is, there can be many global structures, and structures that could be very different, yet yield the same local experience.

The start example is the ol' "you could be inside a black hole an not even know it" space-time. There is a "null dust" collapsing about you at the speed of light (this is slightly more probably than the infinite mass sheet, but still highly unlikely, of course :) ). You are sitting there floating in locally Minkowski space-time, but it's already too late. The null dust has collapsed below the Schwarszchild radius and there is no escape for you, but you won't know until poof! you're gone.

Put that in your pipe and smoke it. Even assuming GR is absolutely correct, there are things we can't know until we know them. Boundaries we can't see beyond if an until something catastrophic happens.

Does a singularity lurk behind a black hole horizon is but one of many unanswered questions. Must all matter and energy that crosses that horizon be inexorably bound to end in singularity?

And what's beyond the Cosmological horizon, for instance?

My point is the question of boundaries we can't see beyond is really something different that whether singularities lurk beyond some of those boundaries.

-Richard

Ken G
2008-Oct-17, 05:35 AM
THe Rindler horizon is also a boundary an accelerating observer can't see beyond. You're on a space-ship accelerating and you have been since forever. There's a big black horizon behind. A "infinite mass sheet" could lurk behind there. Or nothing could. :lol:

Is this the same question? And it's a matter of coordinates. But what's different about black holes in that regard? I think there is a difference. The Rindler horizon is a kind of fiction, because it does presuppose the acceleration has been ongoing forever. For temporary accelerations, it is merely the distance behind me that time appears to stop, and further behind there, time reverses direction. So that's purely a coordinate thing-- I can only not get empirical data from across a Rindler horizon if I maintain my acceleration. The black hole event horizon seems a lot more "real" because there's nothing I can do to receive that information, it's something reality is doing rather than something I'm doing. Maybe that has something to do with the different spacetime curvature involved, and the fact that a black hole event horizon interchanges the radial coordinate with time, whereas I think the Rindler event horizon only reverses the direction of time.

There is the point that a local space-time, an observable space-time, can't tell you the global space-time. That is, there can be many global structures, and structures that could be very different, yet yield the same local experience. Yes, that's reminiscent of the "image charge" problem.

Put that in your pipe and smoke it. Even assuming GR is absolutely correct, there are things we can't know until we know them. Puff, puff. So the next question there is, does this mean that relativity provides an incomplete ontology, or is that the only ontology that has meaning-- the ontology that extends not one iota past what you can know (the epistemology). I tend to argue the latter in the "Heisenberg cut" kinds of situations, so I suppose that's how I'd lean in relativity too.


Boundaries we can't see beyond if an until something catastrophic happens. Yes, but there seem to be two types of such boundaries-- some that are always opening and closing, like the way if you run wind sprints you can make time go backward in the Andromeda galaxy (using your own global concept of "now", that is), and those that seem more permanently encoded into spacetime through no fault of your own, like event horizons.

And what's beyond the Cosmological horizon, for instance? Yes, that's the one I'd see as analogous to a black hole event horizon, moreso than I would a Rindler horizon. It's not clear, though, how much of this is ontologically significant, and how much is pure coordinates.


My point is the question of boundaries we can't see beyond is really something different that whether singularities lurk beyond some of those boundaries.Yes, I'm not even looking at singularities, which we kind of know are unphysical, I'm just interested in "what's behind the door" in any sense. It is weird enough, for example, that the radial coordinate inside a black hole event horizon turns into a time coordinate, for example-- notwithstanding any singularity issues.

Ken G
2008-Oct-17, 05:47 AM
I think that the notion that matter is fundamentally composed of atoms is pretty well accepted. There is pretty good evidence that it is true.I'd say that really depends on what you mean by "fundamentally composed". If you just mean, "it affords us with a whole lot of unifying and organizing power to imagine that reality is composed of atoms", then I can agree completely. But that doesn't necessarily sound like "fundamentally composed". To get a sense of the disconnect, I would happily wager that in another millennium, if humans have not done anything awful to themselves, their science will be such that they will look back at the concept of "atom" as a naive notion of our current time.

Note we already do that with the Greek notion of "atom"-- an atom is supposed to be indivisible, but we kept the term even after we found we could split them. But even if discreteness rather than indivisibility is the key attribute to which you refer, it is still true that if the LHC finds the Higgs boson, we will go around saying that particles are afforded their mass by spontaneous symmetry breaking in the Higgs field, or some such thing, which will not sound much at all like "matter is fundamentally composed of atoms". It'll sound more like "matter is fundamentally composed of abstract mathematical goo of some kind, but the less sophisticated can imagine that means atoms".


But the evidence is basically the fact that models and predictions based on the atomic model work exceedingly well. That is the right way to say it, that's what science tells us. Whether or not it tells us anything more than that is where we find the murky terrain.


Similarly generally relativity tells me that black holes exist and describes space-time inside the event horizon (as long as I stay away from the singularity).But see how different that sounds than "predictions based on GR work extremely well"? Why do we feel we need to say anything more than just that which science can actually support? Note that all the predictions we are talking about here apply on our side of the event horizon, as nothing on the other side has been tested. Recall the thorny lesson of Newtonian mechanics and the speed of light.


So, you can either believe your physical models, with the appropriate skepticism of a scientist who knows that all of our physical laws are but good approximations, or you can take the position that only that that I have actually seen and measured is real. Are there other options that have certain advantages over both of those two?


For the event horizon question I suppose that there is one other viable position -- Who cares, since it doesn't affect me? But that rather rejects curiosity which ought to be an anathema to either a philosopher or a physicist.I think that's actually a valid question, I'm not sure we can dismiss it as anti-curious. It may be important to recognize that science must be fundamentally about what does affect us. I think a physicist or philosopher can adopt that stance without being untrue to either art, they just have to be willing to follow it to its logical conclusion. That's all we can ever really do, no matter what stance we adopt.

publius
2008-Oct-17, 06:08 AM
Yes, I'm not even looking at singularities, which we kind of know are unphysical, I'm just interested in "what's behind the door" in any sense. It is weird enough, for example, that the radial coordinate inside a black hole event horizon turns into a time coordinate, for example-- notwithstanding any singularity issues.

The radius/time flip brings up something I almost posted above, but it was getting long. There is an aspect of what happens behind the door of the black hole that is asking "what happens in the future"? In any external coordinates, events inside never happen, they don't map into any external sense of now at any finite time. So, at least dropping to an observer centric view, the question is what happens infinitely far into the future.

Time gets mixed into this in ways that can sort of blow your mind. Spatial boundaries for one observer become temporal boundaries for others.

So is this question any different than asking what will happen in the future for anything?

For example, the laws of physics could suddenly change. :lol: Maybe it's a vacuum collapse event, where a false vacuum collapses into a lower energy state where the laws of physics are completely different. Or something happens that's even wilder than that, something not yet conceived.

So we can't be certain of any theory until we've had infinite time to observe everything. There is a boundary in our future for everything. We're continually crossing a barrier as we move through time, and have no idea what lurks behind it until we get there. So far, it's been pretty uneventful in the grand sense of the laws of the universe, but we don't know until we get there.

So I'd say this doesn't apply to just relativity/black holes, but all of physics, even the simplest, most universally accepted principles. There is a time boundary on everything. Every theory is prefaced with "Up until now, and as far as we can see, blah blah....."

-Richard

WaxRubiks
2008-Oct-17, 09:22 AM
is the belief in the event horizon a sort of religion?

Like Ken G, I would agree that talking about the event horizon as being real without that having been proven, isn't doing science.

As a model it's fine, but what if the singularity and the event horizon are inseparable, and that assuming the singularity isn't real, then neither is the event horizon.

I personally think that space warps(appears to get bigger to preserve the measurable speed of light) around collapsing dying star matter in such a way that it never forms an event horizon, and light and information can always escape.


As I asked in other threads; this should be testable with non-blackhole objects as an apparent size increase; if proportional to the time dilation, the Sun should look wider by about 3kilometers for a distant observer,than close up. But no one seems to have an answer whether this is testable, or testable for Jupiter, or testable in some other way.

RussT
2008-Oct-17, 10:51 AM
For example, the laws of physics could suddenly change. Maybe it's a vacuum collapse event, where a false vacuum collapses into a lower energy state where the laws of physics are completely different. Or something happens that's even wilder than that, something not yet conceived.

This is 'close'...just reverse the processes like this....

Maybe it's a vacuum collapse event, where a false vacuum collapses into a lower energy state where the laws of physics are completely different.

This is Our SMBH's 'vacuum collapsing' from a 'low energy state' to the High Energy Gamma Radiation Bursts.

Once you understand that that IS happening, then you can understand how the 'initial conditions' is the Low Energy CMB is the "Constant" "Leaking" "Straight Through" (Lee Smolin/Lisa Randall) to allow the "Vacuum Collapses" to be able to happen...;)

cjameshuff
2008-Oct-17, 01:53 PM
I'd say that really depends on what you mean by "fundamentally composed". If you just mean, "it affords us with a whole lot of unifying and organizing power to imagine that reality is composed of atoms", then I can agree completely. But that doesn't necessarily sound like "fundamentally composed". To get a sense of the disconnect, I would happily wager that in another millennium, if humans have not done anything awful to themselves, their science will be such that they will look back at the concept of "atom" as a naive notion of our current time.

I would not. We are well aware that considering atoms as indivisible objects is as much a simplification as Newtonian mechanics. However, they *are* discrete objects, have not only been imaged but have been manipulated as such, and the situations in which their composite nature becomes important (beyond the participation of electron shells in bonds) are well defined...better defined than the situations where relativistic effects become significant, really.

Ken G
2008-Oct-17, 02:02 PM
So is this question any different than asking what will happen in the future for anything? Or even, is what will happen in the future part of what is happening now, or is it something different than what is happening now? On that matter, I would point out that it is generally thought that physics can, to some degree, predict the future, but in fact, it never does that-- all it ever does is to predict that past which has not yet happened. In other words, we never test a theory by comparing its predictions to observations that haven't been done yet, we compare its predictions to observations that we just did. If we are talking about a future that, in a sense, will never be the past, how do we do physics on that? Your point speaks to the deep disconnect between what we call the act of doing physics, and the analysis of event horizons.

Or something happens that's even wilder than that, something not yet conceived.I believe The Bard had something to say on that, Horatio.


So far, it's been pretty uneventful in the grand sense of the laws of the universe, but we don't know until we get there. I was just thinking recently along similar lines, which motivates my new signature. Your way of saying it makes it more blindingly clear.


So I'd say this doesn't apply to just relativity/black holes, but all of physics, even the simplest, most universally accepted principles. There is a time boundary on everything. Every theory is prefaced with "Up until now, and as far as we can see, blah blah....."Yes, I think that's exactly right. We just get tired of saying that every time, so we don't-- and then we forget it's true.

Nereid
2008-Oct-17, 02:28 PM
Great posts lately! :) :clap: They are covering some of the things I'd hoped we'd discuss in this thread*.

Just a couple of things quickly (without quoting anyone, for now):

+ if you add certain (other) parts of physics to GR you get Hawking radiation; that opens a window to testing 'reality' (in the scientific sense Ken G summarised in the excellent post that is now in a disappeared thread; rcglinsk's collateral damage?), or at least something about event horizons and black holes; extend things a bit more and you get the scenario tdvance mentioned

+ 'seeing' (DrRocket): no one has 'seen' an atom, and never will ... unless chains of logic and math of exquisite beauty and intricacy (among other things) can be conflated with 'see' (and let's not get started on 'real'!); what allows us to conflate with such confidence?

+ ditto, plus 'lab/experiment': has H.E.S.S. (or MAGIC?) 'seen' that Markarian 501 is a point source of (10) TeV gammas?

+ will anyone ever 'see' a quark?

+ 'shut up and calculate': who cares about trying to wrap their minds around event horizons, singularities, etc? GR (or physics, in general) lets you make testable predictions, isn't that enough? Of course, if thinking about one-way tickets, (partially) partitioned universes, or whatever helps you come up with better predictions (or more quickly or in a way that is easier to communicate or ...), go for it!

+ (blowing my own horn): the 'laws of physics' are everywhere and everywhen the same ... as an axiom don't you think it's a rather handy way of avoiding a great deal of philosophising?

* no sign of 37.1101 though; no idea why, having asked the question (sorta), she has left the building ...

Nereid
2008-Oct-17, 06:52 PM
is the belief in the event horizon a sort of religion?

Like Ken G, I would agree that talking about the event horizon as being real without that having been proven, isn't doing science.
I'm not sure Ken G said that ... if for no other reason than he'd never use the word "proven" in that way (without clear qualification)!


As a model it's fine, but what if the singularity and the event horizon are inseparable, and that assuming the singularity isn't real, then neither is the event horizon.
Yeah, but ...

What else is there in (modern) science than "models", in the sense of your post? (not counting things that aren't as concrete, 'pre-models', or ideas, perhaps)

In the context of this thread (crudely, 'assume GR'), the singularity and the event horizon are inseparable.

However, I think you have the logic reversed, at least if we're approaching this as science: 'the event horizon' is pretty darn testable! And if you read some of the material Tim Thompson posted earlier, you'll see just how close we're getting to doing some good tests (to 'prove' it's 'real'!). Also, 'the singularity' cannot be 'real' in any testable fashion ... except via GR, observing event horizons, or by discovering laws of physics that go beyond (in the appropriate sense) GR.



I personally think that space warps(appears to get bigger to preserve the measurable speed of light) around collapsing dying star matter in such a way that it never forms an event horizon, and light and information can always escape.
Then that would involve one (or more) of those "competing theories" Ken G mentioned, and that'd take us outside the scope of this thread.



As I asked in other threads; this should be testable with non-blackhole objects as an apparent size increase; if proportional to the time dilation, the Sun should look wider by about 3kilometers for a distant observer,than close up. But no one seems to have an answer whether this is testable, or testable for Jupiter, or testable in some other way.
You might like to read up on some of the work done on accretion disks around SMBHs; I think you'll find these sorts of effects have already been modelled, and that some planned observations may be able to detect them (again, see some of the material in Tim Thompson's post, or papers cited in that material).

WaxRubiks
2008-Oct-17, 09:16 PM
I'm not sure Ken G said that ... if for no other reason than he'd never use the word "proven" in that way (without clear qualification)!


yes, sorry; I meant I was agreeing to his general statement about needing actual physical evidence for things in order to do science.



It seems to me that a coordinate system should have as one of its foundations, that light is always to be measured as having the same speed, where ever it is measured from. And to have an event horizon where light can't even be measured beyond would be part of an incomplete model of reality.
I'm sure that the model can produce useful results, but I wonder if those results are the product of a sort of bendy theoretical telescope that goes inside the event horizon and then out again to look at what is going on outside the event horizon.....if you understand what I mean....

It seems to me that the singularity/event horizon model may produce a sort of mathematical limit to what can go on, but that the physical universe just stays away from that limit, so that an event horizon or a singularity never actually form.

Ken G
2008-Oct-18, 01:22 AM
I would not. We are well aware that considering atoms as indivisible objects is as much a simplification as Newtonian mechanics. However, they *are* discrete objects, have not only been imaged but have been manipulated as such, and the situations in which their composite nature becomes important (beyond the participation of electron shells in bonds) are well defined...better defined than the situations where relativistic effects become significant, really.Actually, it may surprise you to learn that this simply isn't true. It is very dangerous to apply ontology (theories of existence) blindly! Yes we get away with imagining that atoms are "discrete" in most situations, but in fact they are not discrete, not in our best understanding of a realistic ontology of the universe. In that theory, all hydrogen atoms, for example, are identical and therefore indistinguishable. This further means that is you write the "wave function" of a collection of hydrogen atoms, you cannot (in principle) write it as a sum of discrete wave functions-- you have to write it as sum over all combinations of exchanging these identical and indistinguishable objects.

Now, in very few practical situations would we ever actually bother to do that, as there would no need for it. I am merely pointing out the danger in making statements that some things "are" discrete. That is simply not the ontology of our best and most complete theory about atoms, it is an ontology of the "effective" theories we use in most real cases to describe atoms. Of course, I would also argue that all theories are effective theories, but that's for another thread.

publius
2008-Oct-19, 05:34 AM
And there is a barrier in the past as well, a Big Bang of a Barrier. :)

Everything is prefaced with "From the Big Bang until now, as far as we can see....." And I emphasize we're only seeing a chunk of that Big Bang.

T = 0 on the ol' cosmological co-moving clock could indeed be a moment where the future decoupled from the past and started over. And that supposes that terms like "future" and "past" meant anything like they do now back then. :shifty:

We are a little bubble of space-time with stuff in it that has become self-aware and started gazing at its navel. What lies beyond that bubble, which has both spatial and temporal limits is unknown. And it will always be.

Some might protest, what if we come up with some grand theory that "explains" the Big Bang. Say that theory explains why this little bubble of reality came to be perfectly. It makes perfect sense and elegantly answers every question about everything that is observed.

Forget about the fact that even though such a theory might answer all the current questions, it would just pose bigger questions that haven't been thought of yet. First of all, could that theory be falsified? Would there be any way to verify it if we can't see beyond the barrier of the Bang Bang event where whatever clock was "reset" in whatever way?

All it would be is postulating what happens on the other side of a barrier that we can never see. :)

-Richard

Ken G
2008-Oct-19, 06:47 AM
T = 0 on the ol' cosmological co-moving clock could indeed be a moment where the future decoupled from the past and started over. And that supposes that terms like "future" and "past" meant anything like they do now back then. In fact, I've heard it said that Hawking and some others imagine that the very early universe was 4-D Euclidean with no time at all, but the universe "began" when time broke off from space and started obeying Minkowski geometry (which is what turned it into time instead of space). I have no idea why he sees it that way, but it just goes to show the kind of bizarre ontologies that the Big Bang model admits.


Forget about the fact that even though such a theory might answer all the current questions, it would just pose bigger questions that haven't been thought of yet. Like the doozy, if we can explain where our universe came from using the laws of that uber-universe, how do we explain the uber-universe? What constitutes an explanation, do we ever reach a point where we can say "OK, well that pretty much puts that issue to bed, now we understand what the universe is"? And people get on philosophy's case for not answering the "big questions" definitively-- I think we imagine all kinds of fantasies about what our intellect is actually capable of establishing. There are two kinds of delusions, those that don't work, and those that do.