Back from New Orleans, and there are now three books I’ve read recently that I’ll try and write reviews of. The first is The Little Book of String Theory, by Princeton’s Steve Gubser. The author has a web-site for the book here, and the introduction is available.
While trying to cover a huge amount of complicated material, the book is quite short, with 162 pages of text, in a small format. Gubser has chosen to deal in a radical manner with the problem of deciding whose work to reference, and whose name to mention in connection with various discoveries. There are no footnotes or end-notes, no bibliography of any kind, and no mention of the names of any string theorists, or any living physicists at all for that matter. The history of the subject pretty much only appears in a few of Gubser’s comments about early parts of his own career.
To somehow counterbalance its main focus on a highly sketchy treatment of an intricate and very abstract subject, the book periodically introduces some very concrete and explicit numerical computations, starting with a first chapter devoted to explaining in detail the equation E=mc2. Unfortunately, none of these calculations have anything at all to do with the topic of the book, string theory. The central section of the book, about branes and duality, contains no such concrete calculations, but instead largely consists of page-long paragraphs recounting in words the intricate structures that occur in this subject. I find it hard to believe that anyone not already familiar with this topic will get much out of this kind of discussion.
Gubser intensively uses analogy to try and convey some understanding of the material, and has a fondness for analogies based on his mountain-climbing experience. Here’s an example, based on a climb to the Aiguille du Midi:
The ridge we climbed is famously narrow, heavily trafficked, and snow-covered. For some reason everyone seems to climb it roped up. I’ve never quite approved of the practice of climbing roped when no one is tied to a solid anchor. If one person falls, it’s hard for the others to avoid being pulled off their feet. Usually I think it’s better to trust yourself and climb unroped, or else anchor and belay. But I’ll admit that I climbed the ridge roped up to my climbing partner like everyone else. My partner was a very solid climber, and the ridge isn’t really that tough.
In retrospect, I think that roped teams climbing a narrow ridge provide a good analogy to the Higgs boson, which is one of the things LHC experimentalists hope to discover.
The point here is that the top of the ridge is supposed to be like the unstable maximum at zero of the Higgs potential, but it seems to me that few are likely to get much real understanding out of this kind of analogy. Similarly, the chapter on GR and black holes opens with a chilling story about a fall while climbing near Aspen, but it’s hard to see how it adds much to the reader’s understanding of the subtleties of the modern understanding of gravitation.
The book is advertised as “a non-technical account of string theory and its applications to collider physics.” The last chapter is about recent attempts to use AdS/CFT as an approximate calculational technique in heavy-ion physics. This is Gubser’s specialty, and he does a good job of giving a hype-free explanation of the state of the subject, for instance:
The second reason why it is tricky to compare a prediction of the gauge/string duality with data is that the string theory computations apply to a theory that is only similar to QCD, not to QCD itself. The theorist has to make some translation between one and the other before he or she has a definite prediction to give an experimentalist. In other words, there’s some fudge. The best attempts to handle this translation honestly lead to predictions for the charm quark’s stopping distance that are either in approximate agreement with data, or perhaps as much as a factor of 2 smaller. A similar comparison can be made for viscosity, and the upshot is that the gauge/string duality produces a result that is either in approximate agreement with data, or perhaps a factor of 2 away from agreement.
While giving a reasonable account of the heavy-ion collision story, the description of the relation of string theory to the much more interesting question of what happens in proton-proton collisions at the Tevatron or LHC energy frontier is actively misleading hype. What he is really describing is supersymmetry, and while he begins with the arguable:
Supersymmetry predicts many other particles, and if they are discovered, it would be clear evidence that string theory is on the right track.
he then goes on to claim that:
What is exciting is that string theorists are placing their bets, along with theorists of other stripes, and holding their breaths for experimental discoveries that may vindicate or shatter their hopes…
If it [evidence for supersymmetry] is found, many of us would take it as confirmation of superstring theory
There’s no discussion of the issue of the supersymmetry breaking scale, or acknowledgement of the fact that string theory does not at all require this scale to be low enough for superpartners to be observable at LHC energies. The fact of the matter is that string theory makes no predictions at all about what the LHC will see, and Gubser’s claim that string theorists have some sort of LHC prediction they are betting on is just not true. There is no bet here that string theorists can possibly lose: if superpartners are found, they are likely to trumpet this as “confirmation of string theory”, but if not, they’ll fall back on the accurate statement that string theory predicted nothing about this.
Throughout the book, Gubser is on the defensive about the issue of string theory’s lack of predictivity, invoking highly strained and dubious analogies as excuses. One chapter begins with a discussion of Roman history and its effects on our present-day culture. He then argues that our many centuries remove from this history is somehow like the way string theory makes predictions at high energies, not low energies. I don’t see the analogy (we have lots of evidence for Romans, none for strings), and in any case the problem with string theory is not that it can’t predict what happens at low energies, but that it can’t predict anything at any energy. In another chapter he compares current string theory unification models to the BCS theory of superconductivity, noting that the BCS theory doesn’t work for high-temperature superconductivity. I’m not sure what to make of this analogy, since BCS is a successful theory, string unification models aren’t. The only point of it seems to be the hope that something new will be discovered experimentally (analog of high temperature superconductivity), and some unknown version of string theory will describe it.
Like pretty much all of his colleagues at Princeton, one thing Gubser wants nothing to do with is the multiverse and the anthropic string theory landscape. While he explains the moduli-stabilization problem, the landscape and the multiverse are not discussed, and anthropic argumentation is dismissed with:
Altogether, I find myself unconvinced that this line of argument is useful in string theory.
In the next posting, I’ll write about another new popular physics book, one that I think is much better and much more readable, although it takes the West Coast multiverse interpretation of string theory as gospel, ignoring the views of Gubser and his Princeton colleagues.