Wednesday, June 20, 2007

Tanmay Vachaspati: black stars & there are no black holes

A report on this blog about George Chapline's colloquium remains the #1 hit if you Google seach for black holes don't exist. It is a rather popular albeit untrue sentence that many users want to see. ;-) George is a friend of The Reference Frame but his black hole ideas unfortunately don't make any sense.

In the article under the previous link, we have explained that why the event horizon - a "red" hypersurface in space defined as the boundary separating a causally disconnected region of spacetime (the dark blue triangle, the interior of the black hole) - is created long before the black hole reaches equilibrium and long before it starts to evaporate.

In fact, the place where the horizon appears for the first time (the lower portion of the red line) looks completely ordinary and the people who live there don't have any tools to figure out whether they are already behind the horizon or not: they think that they are inside an ordinary star. If they had such tools, they would be extremely non-local, required superluminal propagation of signals, or a time machine. At that moment - when local physics still looks completely ordinary - it is already decided that the future curvature of spacetime creates a causally disconnected region because the evolution of spacetime according to Einstein's equations will inevitably lead to a spacetime whose causal diagram is depicted above: a spacetime with a causally disconnected black hole interior.

We can also prove that no plausible modification of Einstein's equations that keeps them consistent with observations can remove the conclusion about the creation of the event horizon.

This is a rather trivial conclusion of classical general relativity that most students with an A from general relativity will be able to make. It is extremely robust and whether or not we can test it "directly" is secondary. Equations of general relativity have been verified in other but related experiments and everything else we need a solid calculation that is actually more reliable than the experiments even though many people who are not quite sure about the consistency of mathematics and its relevance for the real world have irrational problems with this statement. ;-)

I always wonder whether the people who don't trust mathematical derivations believe that they would get killed if they jumped from a skyscraper. Have they ever experimentally tested this assertion?

New Scientist

Nude Socialist has just promoted a theory of Tanmay Vachaspati from Case Western Reserve University, Ohio. In his 2-page paper, he argues that black holes are never formed in the first place. Instead, the collapse stops in a stage that he calls a "black star" and he even proposes that a new, non-black-hole-like kind of collision of these black stars is responsible for gamma ray bursts.

Now, I find notion that quantum gravity regulates black holes as something that looks like a black hole but is microscopically just another "regular" object - something I called "not quite black holes" - to be a legitimate paradigm. But of course I know why all qualitative conclusions about black hole dynamics will continue to hold when the classical approximation of general relativity is valid i.e. whenever the black holes are large.

What we know today goes well beyond the classical approximation of black holes. We can calculate the entropy of large classes of black holes arbitrarily accurately, among other things. We simply know that these things are correct. It's a matter of doing the math right.

It is not hard to read the whole Vachaspati's paper and it obviously makes no sense. Does he assume some novel quantum gravity effects? No, he is just talking about classical general relativity. We can easily show that event horizons are inevitably formed in this picture. It is a straightforward exercise for those who know the technology. Of course, it is an uncertain, mysterious sea of dragons for those who don't.

We also observe (the effects of) black holes in telescopes - for example one at our galactic center whose mass counts in millions of solar masses - but I won't hide that the theoretical derivation of their existence from other experimentally known data seems even more robust than the direct observations to me (and others).

Vachaspati - whose list of former co-authors includes Mark Trodden and Lawrence Krauss, among others - rejects this result except that he doesn't seem to give a glimpse of an argument that the result should be different - except for saying other things that obviously don't occur such as a new kind of "pre-Hawking superfast radiation". Moreover, the only figure that is included in his short paper is a standard Penrose diagram for a Schwarzschild black hole. It doesn't look like he has gotten rid of the horizon. Quite on the contrary. It is quite nicely seen on the figure: it's the Northwestern diagonal line.

Nevertheless, this paper was accepted for Physical Review D which is why Nude Socialist happily describes it as science: certain papers are simply vastly more interesting for the journalists than others. Moreover, the Nude Socialist formulates the article in such a way that Vachaspati's weird paper must surely be very important and 't Hooft and Giddings are just frozen ultraconservative frogs who inhibit the "progress". Thousands of readers will buy it. New Scientist presents Vachaspati as a hero and for thousands of stupid readers, it's simply enough to become convinced.

Most articles about Vachaspati are located in Indian media which is no coincidence.

Also, in his CCNET, Benny Peiser gave it a title "Black hole denier: another scientific consensus in trouble". Benny is a smart Gentleman but it would be dishonest not to say that his title is significantly less smart. If the real driving force of his climate skepticism were a general tendency to fight against anything that others think regardless of the existence of a rational reason, I couldn't agree with him.

Classical general relativity is a settled theory and it is extremely difficult and probably impossible to invent a description that would - at least barely - agree with the same experimental tests but that would be able to stop event horizons from forming. Also, quantum effects can be neglected in the case of large black holes and the alternative black hole physicists even seem to agree with this conclusion.

There is no scientific consensus about the existence of black holes. The people who understand general relativity and its justification know that black holes must exist while those who don't understand general relativity don't know whether there are black holes and most of them probably think that black holes don't exist, at least in the privacy of their homes. I guess that the second group includes a majority of the scientific community. This has nothing to do with consensus, it is about knowledge, talent, and expertise. So I would still prefer the adjective "ignorant" instead of "denier" for any person who studies gravitational physics but is unable to make this simple conclusion even in 2007. It would be foolish to demonize such people because ignorance is the primordial state of affairs.

And that's the memo.


  1. Hi. I am in the category of those who don't understand general relativity and don't know whether there are black holes think that black holes don't exist. I found Vachaspati's paper to be coherent given my limited knowledge. Although I assume your knowledge of physics is vastly superior to my own, I find your counterarguments somewhat unsatisfying, and they don't seem to address arguments actually made in the paper.

    For example, Vachaspati first demonstrates that the event horizon doesn't form in finite time using pure general relativity in the frame of reference of the observer, to which you respond by pointing out that quantum mechanics can be ignored for large black holes. Vachaspati only brings in quantum mechanics in order to demonstrate that the addition of quantum mechanics does not contradict his initial purely classical argument.

    It's trivial to demonstrate with classical general relativity that, to an observer watching from the outside, black holes do not form.

    Either of two complementary arguments are sufficient (both from the point of view of an external observer such as on Earth).

    1. Time dilation approaches infinity at the event horizon as it (almost) forms, thus it does not form in finite time.

    2. Space dilation approaches infinity at the event horizon as it forms, thus the distance to the (forming) event horizon approaches infinity. This argument is so counter-intuitive that only someone who understands general relativity can visualize it.

    As for physical evidence, I am aware of the existing evidence, but I see no reason why we would be able to distinguish between a real black hole and one of Vachaspati's black stars. Both have nearly identical light bending properties, and both do a pretty good job of preventing light from escaping.

    Since the argument is simple and compelling, I would expect the counterargument should also be somewhat comprehensible to mere mortals such as myself. Can you provide us with or refer us to a real counterargument.

  2. wolk wrote: You, the observer sitting at safe distance would very shortly observe the very last photon emitted from the infalling body, how that would not mean information loss?

    If you put the body in a bag and buried it where nobody would find it, would that be information loss? :)

    More seriously, as long as the infalling body does not reach the event horizon (and it never will), it has the potential to issue a (probably futile) SOS message. It would take a very long time for that red-shifted message to traverse the newly created space, but it would eventually reach the external observer, hence no information loss.