Sunday, June 03, 2007 ... Deutsch/Español/Related posts from blogosphere

Alan Guth vs Neil Turok: audio

A radio station in Boston aired a

Neil Turok is implicitly promoting their new book with Paul Steinhardt, The Endless Universe, about the cyclic and ekpyrotic cosmologies. The format how he speaks sounds fine but the content is problematic. The way how defends his theories doesn't quite sound scientific to me. For example, he says that it was assumed that there had to be a Big Bang and right after the beginning, many quantities had to take extreme values. And they question this assumption of one Big Bang - which is already supposed to make their case strong.

I don't think that questioning of assumptions makes someone's framework justified. Questioning of assumptions is only a feature of an approach to physics, not an outcome in physics. In this particular case, a very small Universe at the beginning is not an arbitrary assumption but the most obvious and straightforward implication of a previous successful theory of an expanding Universe, something that has been around at least since Hubble's observations.

Theories are only supported by actual evidence. A theory with an assumption that reproduces data and solves demonstrable problems with previous theories is better than a theory without an assumption that however doesn't reproduce data and only solves its own philosophical problems as opposed to real problems with the previous theories.

I personally don't think that the postulate of many bangs or a complicated evolution of the Universe at the trillion-year timescale solves some real problems per se. It is an artificial, a priori unnecessary addition to physics. Predicting a more complicated or a less complicated behavior at the trillion-year timescale is not an argument in favor of a theory. And if the theory predicts such complicated events just for the sake of it, i.e. if they are just an assumption, they become an unnecessary superstructure that should be cut by Occam's razor.

Turok's statement that their cyclic or ekpyrotic Universes are made more likely or necessary by string theory also sounds wrong to me, especially because of the natural way in which inflation fits into string theory.

Alan Guth has to clarify various things - namely that it is not true that everyone has always believed that time has started with the Big Bang. This was just a classical picture of the evolution but the quantum picture has always been more complex. It is more complex in inflation, too. He interprets inflation as the mechanism that makes the Universe large and smooth while he doesn't try to extend the validity of this picture beyond its realm. That's what I think is a rational approach to these questions.

Someone who claims to have solved the inherently conceptual questions of time must be able to answer all the deep questions of quantum cosmology that are often associated with the far future of string theory. Cyclic cosmology doesn't answer them. It is another theory at the same level of complexity and completeness as other effective theories and it shouldn't be sold as something more than that.

According to Guth, the most serious problem with the cyclic models is a lacking description of the brane collision. Turok thankfully confirms that the existence and nature of the bounce is not a well-established hypothesis. The host unsuccessfully tries to ignite a bloody battle, asking Guth whether the Steinhardt-Turok theory is just gibberish as his previous criticism suggests. Guth answers that it is not gibberish but a conceivable hypothesis whose motivation is lacking. Very true.

Turok thinks that the eternal inflation stands on more mathematically shaky grounds than bouncing branes. Well, it may be the case but if it is the maximum Turok can say, then he shouldn't sell the ekpyrotic Universe as an alternative to inflation because inflation doesn't imply eternal inflation.

Turok says that their theory explains the origin of time and because others don't, they must be wrong. This is a doubly incorrect argument. First of all, it is not true that their theory explains the origin of time. It just makes the origin of time more complicated but the very origin remains as unexplained as in simpler approaches. Second, it is simply not true that a theory meant to explain the broad structure of the Universe at cosmological length scales and its flatness has to explain the origin of time. A priori, these are two different classes of questions.

Many more comments by Turok sound like complete nonsense to me. For example, one of the arguments for their theory is that the "conventional" framework is in crisis because it has led to the landscape of possibilities etc.

Such an argument is irrational because their framework doesn't solve any of the real problems that (partially) justify the existence of the landscape within the standard framework. More concretely, their picture doesn't say anything about the values of parameters of low-energy effective field theory - couplings and the cosmological constant. So how can he mention the landscape as an argument for his picture? If he had a different solution to the cosmological constant problem, that would be different. But he doesn't.

This kind of discussion sounds unacceptable to me. For example, an Iranian "alternative physicist" came to my office on Friday and argued that general relativity and the existence of Dirac's monopoles were flawed and as soon as I started to explain him his errors, he interrupted me by mentioning that Lee Smolin has criticized string theory. I told him that his criticisms of Einstein's, Maxwell's, or Dirac's papers are separated from a contemporary articulate crackpot's criticism of string theory at least by 70 years. They have nothing to do with each other and if his argument is based on such a link, it follows that the argument is bunk.

It is often very difficult for some people to figure out what assertion they have actually proved or partially proved and I am afraid that this sentence applies, to a lesser extent, to Neil Turok, too.

Another bad thing that Neil Turok partially shares is a huge oversimplification of the situation that some people - journalists and those who are close to them - like to make. A listener asks whether an experiment will settle which of these cosmological models is right. Turok answers that the truth will be decided by the (non)observation of gravity waves.

Guth had to honestly explain that the situation is more subtle. The gravity waves predicted by inflation can be undetectably small - and other theories may perhaps be modified to predict the waves, too. Of course that laymen want an extremely clear criterion to find the right answer - something that they could easily understand. But it's simply not true that the laymen may understand everything to get fully qualified answers because the qualified decisions often rely on the knowledge contained in thousands of papers that only the experts can apply properly. For example, if a layman thinks that he understood the essential questions of future high-energy physics from a blue and black crackpot books, then I assure her she hasn't understood anything at all and she's been just manipulated by two very stupid texts.

The host proposed a link between inflation and Christianity, and cyclic Universe and Hinduism. Turok tried to deny this link. Thankfully, he admitted that the Hinduist beliefs don't provide them with extra evidence. Guth agrees that we may know the poetry but we can't be directly guided by it. Both cosmologists are fortunately very different from Lee Smolin who is directly guided by Leibniz's beliefs, among other silly things. ;-)

Janna Levin's comments during the show sounded sensible to me, although they were not terribly revolutionary.

Add to Digg this Add to reddit

snail feedback (0) :

(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o), m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m) })(window,document,'script','//','ga'); ga('create', 'UA-1828728-1', 'auto'); ga('send', 'pageview');