Tuesday, March 18, 2008

Problems with black hole remnants

Cosimo Bambi wrote a new paper of the kind that I consider bad physics. He advocates a theory of black hole remnants. More precisely, it is not a real theory but a random conglomerate of phenomenological speculations: it shouldn't have appeared on hep-th because the "th" factor is completely missing.

I want to use the paper as an example to show certain irrational and unscientific tendencies that are fashionable these days. But I have to start with the technical issue of the black hole information loss, a topic that has been discussed several times on this blog.

Thirty years ago, Stephen Hawking showed that black holes should emit a precisely thermal radiation - a direct consequence of causality and methods of quantum field theory. Because thermal radiation carries no information (it is described by a universal mixed state), the information about the initial state seemed to be lost. The unitarity, a crucial principle of quantum mechanical evolution, seemed to be violated, at least in the semiclassical approximation.

There were three main possible answers:
  1. the evolution is indeed non-unitary; the postulates of quantum mechanics have to be weakened and modified
  2. the radiation is not exactly thermal and carries the information
  3. the radiation is thermal but the information is preserved in a long-lived "remnant" that doesn't disappear

In 2008, every good physicist knows that the second answer is correct. We have explicit systems of equations - in AdS/CFT, Matrix theory, and elsewhere in the theory of quantum gravity that continues to be called "string theory" - that exactly agree with unitarity (and manifestly preserve the information) but that also exhibit all the required quantum gravity processes to be identified with black holes in general relativity. And these string-theory-based pictures agree with the Bekenstein-Hawking predictions of the thermodynamic quantities, too.

Even if you didn't believe that string theory describes the real world, the triumphs mentioned in the previous paragraph actually settle the question. Why? Because they show that Hawking's semiclassical predictions that predicted the information loss are simply not rigorous. They are not necessary. In a complete treatment, the information is preserved even though the complete treatment agrees with the semiclassical approximation wherever it should.

It means that there exist at least N backgrounds of quantum gravity where the generalization of the information-loss arguments is violated beyond the semiclassical level and where the information is preserved. In fact, this is far too modest a way to describe the real situation. In reality, all consistent formulations of quantum gravity we know confirm that the Hawking radiation does carry the information about the initial state.

Thirty years ago, when we didn't know about these descriptions of quantum gravity, it was pretty natural and reasonable to believe that a qualitative conclusion of the semiclassical approximation - the information loss - was also true at the exact level. It is pretty natural for approximations to be qualitatively correct. At the same moment, the conclusion couldn't be proved rigorously. But there was a reason to believe that the information was getting lost.

The situation today is very different. We can actually look at these black hole systems in frameworks that go beyond the semiclassical approximation and the answer is that a particular qualitative conclusion of the approximation - the information loss - simply doesn't generalize to the full theory. Because we have a more complete understanding of these issues, it is no longer "equally reasonable" to make this generalization and to assume that the radiation is exactly thermal and that the information is lost.

The real motivation for considering the answers "1" and "3" has evaporated itself. Thirty years ago, the hypothesis that an exact theory happens to agree with the semiclassical processes but it can also encode all the information could have looked as a miracle. That was why people were ready to believe that the information is lost or that there exist many remnants. But the preservation of the information is no longer a miracle in 2008. We have verified that it simply works. The estimated probabilities of the options 1,2,3 has dramatically changed.

We simply know that the information is preserved and the causality is broken by exponentially small effects. You might imagine that the information is really tunneling from the center of the black hole to infinity. Quantum tunneling is a real effect in quantum mechanics and it can generate new exponentially suppressed effects. They are small but sufficient to restore the full information.

The answers 1,3 have become completely unmotivated and bizarre and the probability that one of them is correct is much smaller than 10^{-10}. So if all people in the world understood current physics, these approaches wouldn't deserve a single person who would study them and promote them. Once we looked properly, we saw that unitarity remains a universal principle of physics and macroscopic causality holds, too.

It was just incorrect to assume that causality holds 100% even in the presence of event horizons. In reality, tunneling induces almost immeasurable, exponentially small corrections and when these small effects are taken into account, the information loss paradox is solved without introducing any new ad hoc concept such as the remnants.

What's wrong with the remnants?

For me, the main wrong aspect of the remnants is that is seems impossible to construct a complete theory that would incorporate the concept of remnants. I am absolutely convinced that this kind of an argument is the strongest one even though it is absolutely disconnected from any need to make new remnant-related experiments or observations.

I can offer you more than some negative reasons to eliminate remnants: in fact, we have a positive argument - a pretty complete theory that not only tells us that there are no remnants but it also teaches us what exactly happens and how. I happen to prefer more complete and coherent systems of answers over incomplete ad hoc answers to subsets of questions.

Various aggressive crackpots have made it fashionable to pay lip service to experimental tests of theories even in cases where it is clear that these tests can't be done. I won't jump on this bandwagon and if you kindly allow me, I will keep on considering various Woits and Smolins who promote this ideology to be irrelevant and cheap anti-scientific simpletons. Careful theoretical arguments remain the most important and the most reliable tools to decide about the validity and probability of various assertions related to quantum gravity and similar advanced scientific questions. Every good expert in quantum gravity knows so and none of them is affected by the populist empiricism.

Incidentally, Sean Carroll recently talked and wrote about the naive populist empiricism and I fully agree with him, including the video with John Horgan (except for Sean's comments on the arrow of time).

Cosimo Bambi is clearly another example of a physicist who thinks that all theories in physics are useless and one should never rely on careful calculations and analyses. So the only problem with remnants he sees are some astrophysical bounds. They're relevant - and in fact, they are sufficient to show that remnants are more or less impossible without using other tools - but they are the layperson's method to try to answer conceptual questions such as the information loss puzzle.

Because Bambi doesn't study any theory in detail and only looks at the roughest available astrophysical data, he arrives to completely different conclusions. Let us play his game for a while and imagine that it is legitimate for a physicist to neglect most of the evidence in this context - and most of the evidence is theoretical in character.

Bambi is led to a model with remnants that are not easily seen and he is thus designing contrived methods to hide them. All these things are done in order to keep an answer to Hawking's question alive - even though all good physicists know that it is no longer alive and we actually know a better, more complete, and more consistent answer. One of the weird features of his models is that the coupling constant for the black hole remnants is

  • g = 10^{-10^{122}}.

That's a damn small number. We have explained, with Arkani-Hamed, Nicolis, and Vafa, that it is arguably forbidden to have tiny coupling constants in a theory of quantum gravity without introducing new and lethal light objects. Of course, our derivation is using assumptions and reasoning that Bambi would probably dismiss because he would dismiss any procedure that relies on the human brain. But what do you think, who is doing a better science here?

The expo-exponentially small coupling constant is probably impossible in the real world but even if you accepted it as a possibility, I think that you have to agree that Bambi's comment about the value below the equation 7 in his paper is deeply irrational. He says:

But we might still hope that such a tiny value can be explained in the full theory of quantum gravity.
Wow. What science is hiding behind this strange sentence about a hope? The 9/11 hijackers might have hoped that they would be given 10^{10^{122}} virgins for their services to Allah. Should science be reduced to hopes and hypes promoting these hopes? The hope seems pretty important for that paper. Should papers be built on audacious hopes? Is this the new kind of science?

Rationally speaking, it is very likely that there exists a universal rule that forbids these super-tiny values of coupling constants completely. But even if you look at this question from the perspective of other frameworks in science, such as effective field theory, you will conclude that the super-tiny coupling constant is virtually impossible. For example, if you use the rules of naturalness, the probability of having a tiny value "g" of a coupling constant is comparable to "g". So the probability that Bambi's hope is viable is comparable to
P = 10^{-10^{122}}.
Once again: that's a damn small number. Why should a scientist hope in these ways? To promote a theory that is unmotivated theoretically, disconnected from the rest of science, almost completely excluded empirically, and that tries to solve a question that has already been solved by a completely natural system of insights?

Is it legitimate for a scientist to take a speculation whose probability is, according to the standard scientific tools to estimate probabilities, expo-exponentially tiny and pretend that it is a sensible alternative that should swallow something like one third of the research? I beg to differ. If a possibility looks extremely unmotivated and unlikely according to the cutting-edge expert criteria, it should get much less attention than the theories that actually make sense, that agree with the world as we observe it and with its generalized principles. One paper about silly models is one too many.

Does Bambi has someone around to tell him that his speculations are scientifically inexpensive, don't respect any standards of rigor or realistic estimates of probabilities, and the main reason why he wants to construct these models is his superficial knowledge of black hole physics and his ignorance about many detailed principles and insights that are essential in this discipline?

I think that similar papers are going to spread if the haters of theoretical physics - those who like populist proclamations about "testability" and similar crap - are not eliminated from the system. A vague and irrelevant connection with a randomly chosen astrophysical observation could become a sufficient reason to promote and publish incoherent papers that don't fit together and that are based on assumptions whose probability is, by a scientific counting, comparable to 10^{-10^{122}}.

The standards in physics began to deteriorate and loads of unintelligent dopes have contaminated the scientific community and its suburbs in the name of political correctness, diversity, promotion of mostly irrelevant links with other groups such as the experimenters (in the context of a purely theoretical discipline such as black hole thermodynamics), and similar crap.

They are using modern channels of communication to strengthen their case - and their goal is exactly the opposite from the very purpose of science which is to eliminate crappy ideas. The main technique of the scientific method is exactly to throw papers such as Bambi's paper to the recycle bin. Instead, the media and blogs invent stories about diversity and victims to promote similar crap. If this tendency is not stopped, it will feed itself because a new generation of physicists without standards and qualification will influence the creation of the following generation.

That would lead to the end of science as we have known it but we can stop the bandwagon before it's too late.

And that's the memo.

No comments:

Post a Comment