Information Salvation

July 16, 2004 8:25 PM

It would seem that black holes may not lose information after all, in which case Stephen Hawking has lost another bet.
posted by Songdog (24 comments total)

Quantum physics asserts that matter has properties constituting information which cannot be destroyed. But twenty-eight years ago Stephen Hawking showed that black holes radiate their matter away in the form of informationless energy. Thus was born the black hole information paradox: black holes must destroy information which cannot be destroyed. In 1997 Stephen Hawking and Kip Thorne made a bet with John Preskill on the ultimate resolution of this supposed paradox. Hawking and Thorne bet that information must be lost in a black hole; Preskill bet that it must not.

Recently, however, string theory work by Samir Mathur and others has suggested a way in which information might escape from a black hole after all. It would seem that Stephen Hawking has now also come up with a theory (or endorsed Mathur's) allowing the escape of information, for he recently requested at the last minute that he be allowed to make a presentation at next week's 17th International Conference on General Relativity and Gravitation in Dublin, in which he will explain his reasoning, after which he and Thorne will owe Preskill an encyclopedia.

This isn't the first time Hawking has raised the stakes of his research.
posted by Songdog at 8:28 PM on July 16, 2004

Um..., Yeah! What Songdog Said! Double... uhhh and his other Music Career was a mistake too..!
posted by Elim at 8:57 PM on July 16, 2004

So that's where all my socks have gone. Think I'll ever get 'em back?

I love stuff like this.
posted by WolfDaddy at 9:19 PM on July 16, 2004

This information has been retrieved from a black hole... see... told ya... heh...
posted by moonbird at 9:38 PM on July 16, 2004

Beautiful work and fascinating reasoning. This is one more gravitation thought experiment that seems to point towards the contention that all of nature is subject to quantum rules, without exception.

If this survives peer review this will be a very important result and continues a long line of triumphs in theoretical consistency for quantum physics. If it stands up then further advances in quantum field theory on the road to the development of a theory of quantum gravity (and presumably an almost immediately subsequently apparent unification theory) now have one less problem to deal with.

Importantly the principle of nonlocality can be extended beyond the event horizon, string theory gets a shot in the arm and inflationary theory likewise gets an indirect boost from the successful incorporation of Mathur's work.

The key phrase is: ‘if this survives peer review’.
posted by snarfodox at 10:01 PM on July 16, 2004

If this survives peer review? I thought this was Hawking giving in to a group of peers who already feel this way.
posted by scarabic at 10:18 PM on July 16, 2004

ooooooooh! new theories automatically beat old theories
posted by Satapher at 10:26 PM on July 16, 2004

scarabic> I thought this was Hawking giving in to a group of peers who already feel this way.

The paradox is his proposal, see Songdog's link. Someone had to fix it, he just happened to do the job himself this time around. To quote Frank Herbert: “Paradox is a pointer telling you to look beyond it. If paradoxes bother you, that betrays your deep desire for absolutes. The relativist treats a paradox merely as interesting, perhaps amusing or even, dreadful thought, educational.”
posted by snarfodox at 10:40 PM on July 16, 2004

I'm not able to keep up with that link right now. I need sleep. I'll take your word for it. But I guess the idea of peer review is kind of moot with highly theoretical subject matter anyway. None of these guys knows why he farts out his ass instead of his nose. They're theorizing. This is controversy. Not a rigorous process subject to publication & peer review.

Besides if Hawking has the guts to prove himself wrong and come forward with the information, he might lose the bet, but who cares? That's dilligence and honesty.
posted by scarabic at 11:30 PM on July 16, 2004

The Euclidean path integral over all topologically trivial metrics can be done by time slicing and so is unitary when analytically continued to the Lorentzian. On the other hand, the path integral over all topologically non-trivial metrics is asymptotically independent of the initial state. Thus the total path integral is unitary and information is not lost in the formation and evaporation of black holes. The way the information gets out seems to be that a true event horizon never forms, just an apparent horizon. -- Abstract for Stephen Hawking's talk at GR17 conference

Some small and indefinite perspication, some flirty and indefinite speculation, and some woebegaunt but titillating linkulation at Preposterous Universe.
posted by Opus Dark at 12:00 AM on July 17, 2004

*shrug* quantum mechanics has been, to date, the most accurate predictive theory in human history. Accusations of naïve empiricism aside, if they're making this stuff up as they're going along they've been remarkably lucky so far.
posted by snarfodox at 12:12 AM on July 17, 2004

In theory, proving something very worthwhile to the scientific community is worth more than ones bet.
posted by Keyser Soze at 12:48 AM on July 17, 2004

It would seem that Stephen Hawking has now also come up with a theory (or endorsed Mathur's) allowing the escape of information...

I would doubt that Hawking's solution is going to look much like Mathur's string theory work, for the simple reason that Hawking has historically had very little to do with string theory. It's conceivable that they agree in some effective-field-theory kind of way, but even if they do it probably won't be evident initially.

I'll be at the GR17 conference mentioned above; I'll try to remember to post a summary here when I get back (or sometime during the week, internet access pending.)
posted by Johnny Assay at 6:17 AM on July 17, 2004

quantum mechanics has been, to date, the most accurate predictive theory in human history.

Do you happen to have any links detailing the predictive successes of quantum mechanics, preferably ones intelligible to laymen?
posted by DaShiv at 9:25 AM on July 17, 2004

Do you happen to have any links detailing the predictive successes of quantum mechanics, preferably ones intelligible to laymen?

100 Years of the Quantum (pdf)
posted by eddydamascene at 11:04 AM on July 17, 2004

Besides if Hawking has the guts to prove himself wrong and come forward with the information, he might lose the bet, but who cares? That's dilligence and honesty.

That's science.
posted by rushmc at 11:26 AM on July 17, 2004

Thanks for that link, eddydamascene! Two particularly relevant excerpts from that pdf for me:

That pragmatic approach proved stunningly successful. Quantum mechanics was instrumental in predicting antimatter, understanding radioactivity (leading to nuclear power), accounting for materials such as semiconductors, explaining superconductivity, and describing interactions such as those between light and matter (leading to the invention of the laser) and of radio waves and nuclei (leading to magnetic resonance imaging).

[...]

According to a recent estimate, about 30% of the U.S. gross national product is now based on inventions made possible by quantum mechanics.

Glad to hear how useful and pragmatic all this stuff is. It's hard to see how right now, but maybe someday this talk about black holes and the nature of singularities will have a similar use for us ordinary folks. I'm still rereading the pdf to figure out what collapsed states are and whatnot...
posted by DaShiv at 4:08 AM on July 18, 2004

Fascinating stuff, and I certainly hope Johnny Assay will post that summary. Thanks, Songdog!
posted by languagehat at 2:48 PM on July 18, 2004

You're welcome! And by the way, it's great to find gravitational and quantum physicists hanging out here at Metafilter. Thanks for the links, all!
posted by Songdog at 2:59 PM on July 18, 2004

hey saw this in sciam the otherday...

SCIENTIFIC AMERICAN: What are the top questions bedeviling physicists today?

LAWRENCE KRAUSS: Three that I find fascinating are: What is the nature of dark energy? How can we reconcile black hole evaporation with quantum mechanics? And, finally, do extra dimensions exist? They are all connected. But someone is going to have to come up with a totally new and remarkable idea.

i wonder if it might be the guy who prooved the poincare conjecture?
posted by kliuless at 9:38 PM on July 19, 2004

Yahoo has a news item on Hawking's actual talk. I look forward to Johnny Assay's report!
posted by Songdog at 9:58 AM on July 21, 2004

Alrighty, I'm back, and as promised, here's a none-too-short account of Hawking's lecture. I'll divide it into two parts: the science, and the media. Disclaimer: the opinions expressed here are my own and not those of my advisor, the Scientific Establishment^TM, most of Kansas, or a herd of tapirs.

First, the science. I should say that "I am only an egg" as far as this particular corner of physics goes, but here's the basic gist of it: One of the ways of interpreting quantum mechanics is to say that a system doesn't take a definite path when it goes from state A to state B; instead, it can be viewed as taking all possible paths, with paths that are the closest to the "classical" path being the most likely. This is called the "path integral" approach; it was pioneered by Feynman, and should be familiar to anyone who's read QED.

So far so good. Hawking's resolution to the information paradox, then, is the following (as far as I can tell): First, what we define as "information" can really only be defined far away from the black hole. This is, I think, a consequence of how quantum field theory works in curved spacetimes, but I could be wrong about that.

Now, in the path integral interpretation, a spacetime must be viewed as taking all possible "paths", i.e., to get a physical result we have to sum over all possible histories of the spacetime. Hawking's contention is that when we try to make measurements far away from the black hole, the contributions to this sum from spacetimes where information is preserved (the "topologically trivial" spacetimes) completely dominate over the contributions from the ones where information is lost (the "topologically non-trivial" spacetimes.) Thus, if you stare in at the black hole from "infinity", you never see information disappear because the thermal radiation that we thought would replace it only has a vanishingly small chance of being observed at infinity.

That's what I think he's saying, at least. Most physicists (string theorists excepted, who are basically saying this proves they were right all along) want to look at the promised scientific paper before making a judgement. I believe that Hawking mentioned it would come out in a month or so, so we'll see what happens then. All the people I talked to about it (with the exception of Gary Horowitz, who was the most string-oriented of anyone I talked to about it) had reactions ranging from "I'll make my decision when the paper comes out" to "the formalism he's using is nonsensical." (This sentiment doesn't refer to the path integral formalism — that's pretty much established — but rather to the Euclidean gravity formalism, which essentially involves treating time like it was a spatial direction and calculating things that way. This has the advantage that it makes the answers well-defined, but has the disadvantage that it may not actually correspond to the real physical case of Lorentzian gravity. OTOH, to the best of my knowledge nobody's managed to come up with a well-defined way to apply path integrals to Lorentzian gravity.) Over at the Preposterous Universe, Sean Carroll (a professor in my department) has posted two representative quotes on the subject. (One of which is from my advisor. It is left as an exercise to the reader to figure out which one.)

Now: the media. It was an interesting experience to have the press descend upon the conference, simultaneously bemusing and annoying. The bemusing part was the amount of hubbub made about what was really a very abstract scientific talk. The hall where the talk was given held about 1000 people, and the side aisles were completely lined with media equipment. Hawking arrived 15 minutes late (which, to be honest, wouldn't have been excusable had there not been a media circus surrounding the whole thing), and although I was on the other side of the room it looked like a celebrity entering the Oscars, with flashbulbs going off on either side of the aisle as he entered the hall. Security had been fairly "tight" throughout the day, meaning that they were actually checking delegate badges and not admitting anybody without one; this may sound like a very simple precaution, but no such precautions were taken during the other days of the conference, and it was kind of annoying to have the generally relaxed, cordial atmosphere at the conference imposed upon in this way.

What was more annoying, though, was how the media treated it. It seemed like the media thought this was some truth being handed down from Cambridge on stone tablets, but that's just not how science is supposed to work. At the very end of the NYT article on the subject, the author mentions that Kip Thorne didn't concede the bet; what are not mentioned are Thorne's very sensible remarks (made during the press conference) on how scientific advances are all founded on skepticism, and that once he gets a chance to look at the forthcoming scientific paper and better understand the arguments presented, he may very well concede the bet as well.

This kind of thing isn't usually a problem with science journalism, since the subjects of such articles (at least in physics) are theories that, while possibly still speculative, have been thought about by many people already and have at least a reasonable number of reasonable adherents. In this case, Hawking's celebrity, the back story of the wager, and perhaps the fact that the summer tends to be slow news-wise, combined to create a very different kind of science article, one that's based on a single person's ideas. Granted, that particular person made some fundamental scientific contributions to his field, but having the media base their stories on something that sounds an awful lot like revealed truth makes us physicists a little uneasy.
posted by Johnny Assay at 7:04 AM on July 26, 2004

Johnny Assay> but that's just not how science is supposed to work.

True, but we've always had celebrities amongst the scientific community, for better or for worse. Einstein's support for Bose is a classic example of a celebrity endorsement and his opposition to Heisenberg (until Bohr countered him) was a great example of the opposite affect.

I don't think this kind of publicity does a lot of harm, and it certainly doesn't impede peer review. Andrew Wiles' 1993 paper about Fermat's Last Theorem got slammed in peer review which was no doubt very unpleasant for him, but he ended up replying in the best way possible by addressing the paper's problems in 1994.
posted by snarfodox at 3:00 AM on July 27, 2004

thanks for the report Johnny Assay! here's john baez take btw; pretty much what you sed :D
posted by kliuless at 8:56 PM on August 1, 2004