Science is messy
I disagree with that completely. And yes, I know very well what I'm talking about.
Well,
there's a convincing scientific argument!
Seriously, mate, everything in that paragraph should be sourcable. Tell me
something specific that you disagree with and I'll see if I can dredge up a supporting quote or reference. At least give me
some sort of evidence or argument to back up your position.
John Preskill 2004:
Preskill 2004 said:
Hawking had precipitated a genuine crisis in fundamental physics, and it seemed that we would have to give up at least one of our cherished beliefs. Hawking's radical suggestion was that the foundations of quantum theory needed to be revised. "
On Hawking's
volte-face:
Preskill 2004 said:
This past year he has been thinking a lot about how his earlier conclusions about information loss might be evaded, and in his talk in Dublin last Wednesday he outlined a new argument supporting the conclusion that information loss does not occur after all. Unfortunately, I don’t understand this argument well enough to attempt to summarize it here. ...
====
Nonsense. Science just doesn't work that way. You don't throw old theories in the trash and start from scratch, ever.
You seem to be making quite a few unverifiable absolutist statements. That's usually a bad sign.
We certainly threw Phlogiston theory away, didn't we? At the time, it was the best theory of chemistry we'd ever had, by a long shot. But wrong. The weight change in reacting compounds wasn't due to the flow of energy after all, it was due to a previously-undiscovered element called oxygen. Beautiful theory, philosophically way ahead of its time ... but wrong. We kept some of its language and conventions, but the theory itself is long gone.
Now, okay, it's unusual for a theory to be as successful as Phlogiston theory was, while being
fundamentally wrong, and perhaps there's a limit to how wrong current theories can be, but there's definitely a margin of potential wrongness there, and the conflict between GR and QM is a dead giveaway: something
has to be wrong with the current picture. Until we know what it is, we won't know for sure how much has to be changed.
Yes. Why? Because the old theories worked really really really well, but when you figured out how to test them extremely precisely, you discovered they weren't quite exactly right. So you replace them with a better theory, but one which includes the old theory as a limit.
Not always. Aether theory (as a broad subject) wasn't a superset of failed Newtonian emission theory, and special relativity wasn't a superset of aether theory. General relativity (if you squint really hard) might be classed as a special-case non-particulate aether model (Einstein 1920), but I doubt if you could find anyone who'd consider it to be a
superset of previous aether models.
SR is a superset of a
cut-down version of Newton's model, and GR1915 is designed to be a superset of SR ... but since SR and GR apply some quite different rules (GR associates energy with curvature, SR describes arbitrarily-great energy-concentrations in the absence of curvature), that's possibly not one of the most impressive aspects of current GR.
Now, okay, each theory's numerical predictions have tended to be an incremental improvement on those of the previous theory, in the areas that we considered to be most important when these transitions happened. But with hindsight, some of the other predictions that we weren't so interested in actually got worse. Ballistic emission theory was used to correctly predict the r=2M horizon surface way back in the C18th, along with gravitational light-bending, and gravitational shifts, but the aether theories that replaced it didn't tend to predict these things - this part of the subject actually regressed. Moving forwards to the C20th, general relativity predicts a zero temperature for gravitational horizons, but if we choose to believe the QM description, nasty old c1800 emission theory (with its positive horizon temperature and indirect radiation effects) arguably gives a better correspondence to QM in this regard than 1950's GR does. So although Einstein's general theory was an advance in most respects, in
this respect it introduced a different step backwards.
Trans-horizon radiation wasn't on our list of important subjects when we drew up GR1915, partly because the theory was originally devised for a horizonless steady-state universe. If it had been designed around the idea of an expanding universe, the subject of leaky cosmological horizons should have come up, and we might have ended up with a different sort of theory. The nature of our current general theory was partly shaped by a series of historical accidents (regarding the timing of different discoveries), which influenced Einstein in the selection of a particular set of design criteria. If history had played out differently, we could have gotten a different theoretical wish-list, and a general theory that made a few different predictions to the current version.
So this idea that we always move forward, and we always incorporate all our previous work in the next model, and our current theories are all inevitably decided by the physical data really doesn't seem to correspond to historical reality. It's a nice story to tell schoolchildren, but out in the real world, humans are messy creatures and our science often develops messily, too. We've found that a certain amount of randomness in the system means that we get stuck less often.
It is totally impossible that general relativity and quantum mechanics are wrong.
No.
QM? I actually like QM, and would put good money on it never being overthrown (I was on the record as being against Hawking's pre-2004 position, when he wanted to alter it) ... but I wouldn't be crass enough to announce that it can
never be wrong. I don't personally see how it
could ever be wrong, but I'm not arrogant enough to proclaim that just because I can't imagine something, it can't be possible (I also prefer to put some distance between myself and the sorts of people who often make those sorts of statements).
GR? Now, GR is more problematic. If you go through the history and the psychology of the subject, and you look at a few of the "special" definitions invoked in the GR textbooks and open them up to be slightly more reasonable, you should see that there's a loophole that allows at least one other class of general theory of relativity to exist that isn't a simple superset of GR1915, and which makes different predictions about information flow across curvature horizons. Someone who's highly trained to expert level
in GR1915 might not be able to see it, but if you go back to first principles and try to rebuild a general theory from scratch without making the usual historical assumptions, you should be able to see the second solution.
Now, I'm not claiming that the second model is right, or that there can't be other alternative models that
I can't see (I don't see how, but ...) ... but it's there. You should be able to use existing textbook definitions to "prove" that that's impossible - I can, too. But the precise wording of some of those definitions sometimes seems to have been selected for no other reason than to produce a snug fit with the
particular sort of general theory that Einstein came up with.
FWIW, there's a general principle that I find useful when judging the reliability of information: the
Titanic Principle. If someone says that a thing is probably true, it probably is. If they insist that it absolutely, definitely, MUST be true, and that for it not to be true is just unthinkable ... then the thing is often wrong.
Horses that "can't lose" and plans that "can't fail" usually do.
They are simply inexact -
No, general relativity's
qualitatively-new prediction about gravitational horizons was that they had a
perfect zero temperature.
That's not just "inexact" ... it's precise, it's qualitatively different to the QM prediction, and (if we believe the QM version) it's qualitatively wrong.
Put another way, the margin of error in temperature, expressed as a proportion between the temperature that that GR predicts and what QM (and just about every other theory) predicts, is
infinity.
If you don't accept a proportional error margin of "infinity percent" as representing a wrong result, then we have different concepts of what "wrong" means in the context of a theory's predictions.
... but at a level that's undetectable right now, and may well remain so for the forseeable future of the human race. While unifying QM with gravity is an extremely interesting topic intellectually, it's hardly a burning issue practically speaking.
Well, if physical effects that have never yet been measured, and might not ever be measured as naturally-occurring events in the forseeable future of the human race aren't interesting, then why the heck did we just spend all that public money building CERN LHC? We did it because we wanted to know what happens in extreme situations that we'd never normally get the chance to see in nature. We want the bigger picture. Research on quantum gravity tries to extract data for other extreme situations, using bright people who can think about this stuff and generate answers without having to spend a few billion of GDP on a dirty great chunk of hardware. I mean, for most people, whether the Higgs boson exists isn't exactly a burning practical issue either.
You can defend LHC by saying "Ah, but the unknown things that we may discover in these extreme realms may have unforseen benefits. They may suggest new achievable technologies that we can't yet imagine, and whose benefits we can't yet quantify." Well, the same goes for quantum gravity.
Sure, we can't visit a natural, solar-mass black hole and measure the Hawking radiation, but we can devise experiments to measure the analogous indirect-radiation effect through horizons in Bose-Einstein condensate, or in experiments in nonlinear optics, and there are a decent number of experimenters working on these things right now, as proper engineering.
And if you don't believe that the conflict between GR and QM has any current practical importance, consider this: If we hadn't done all that theoretical research on Hawking radiation and decided that this particular conflict between GR and QM
has to be settled in favour of QM, then scientists would never have been allowed to build the Large Hadron Collider in Geneva, would they? If we still thought that tiny black holes were immortal, and didn't just fizz themselves away in a burst of Hawking radiation, we'd have to send an international demolition team into Geneva to take that complex apart with explosives. Our willingness to create and operate LHC (without taking a shedload of tranquilisers) demonstrates that three-and-a-bit decades is long enough for us to do a 180-degree reversal over something that we were previously told was a piece of mathematically-proven physics that couldn't
possibly be wrong.
Things change.