Not Even Wrong

Two of the year’s largest particle theory conferences are taking place around now, with Lattice 2004 attracting 280 physicists to Batavia, Illinois this past week and Strings 2004 drawing almost 500 to Paris starting tomorrow. Normally I feel kind of sorry for string theorists since their field is in such bad shape, but this week I’m jealous since I would have loved to have an excuse to go to Paris this summer (I’m not jealous of the lattice gauge theorists who are getting to spend the week in Batavia).

Maybe I’m wrong about this, but the Paris string theory conference seems to me to be the largest gathering of particle theorists that I can remember ever having taken place. In recent years these things have been huge, with attendance around 400-450, but this one should be even larger. It is so over-subscribed that they haven’t been taking on-line registrations for weeks.

Both conferences should have transparencies from the talks available soon on-line and the list of titles and speakers in Paris is now available. Some of the ones that look like they might be interesting are Robbert Dijkgraaf speaking about “Topological M-theory”, Nikita Nekrasov on “Chasing M/F theory” and maybe Greg Moore, whose title is the mystifying “Anomalies, Gauss laws, and Page Charges in M-theory”. One big theme of the conference looks like it will be N=4 super Yang-Mills theory. This is an interesting and well-defined quantum field theory, and one can study it whille claiming to be a string theorist because of the AdS/CFT conjecture and Witten’s recent work reformulating it in terms of topological strings in twistor space.

Next Saturday in Paris there will be a whole day of talks devoted to the unrelenting hyping of string theory to the general public, something which is a standard feature of the “Strings XXXX” conferences, but not the “Lattice XXXX” ones. Somehow I suspect the speakers will neglect to emphasize the utter lack of any progress towards making any contact with reality during the past twenty years. In case the Paris conference is not enough, there are quite a few satellite conferences, including a pre-conference workshop at the IHES and post-conference workshops at CERN and Durham.

Update: It looks like Jacques Distler will be reporting direct from the conference.

Further Update: It seems that they have a WiFi connection at the Paris conference site. From my web server logs, it appears that one thing attendees at the conference are doing during the more boring talks is reading “Not Even Wrong” on the web. Hi Guys!

Just got the following e-mail:

Date: Fri, 25 Jun 2004 19:40:58 +0200
From: fau.alex (fau.alex@wanadoo.fr)
To: woit@cpw.math.columbia.edu
Subject: Mr Woit

hi,
I am a french student and I will go in 1re(in France) the next year
and I want to speak with specialists because I want to work in science
when I will be adult … That’s why I write to you !
I want to know who are you ? What did you study, …
Thanks in advance for your answer
Best regards
Alexandre Faure

At first I thought “Oh yes, just another in the avalanche of fan-mail. Wonder if I should take the time to respond?” Then, looking more carefully at the HTML version of the e-mail, I found an embedded link at the bottom:

href="mailto:liu-yang.imp@th-phys.edu.hk"

That address looks awfully familiar.

The KITP at Santa Barbara seems to be going all out to hype claims of observability by LIGO of effects of cosmic strings. The front page of the KITP website prominently features the story, and adds the personal phone numbers of the authors to encourage the press to contact them (something I’ve never seen theoretical physicists ever do before). Maybe these guys have already hired an agent.

After an expensive upgrade to increase its luminosity, the Tevatron at Fermilab was turned back on in March 2001 to start “Run II”, during which it was hoped that the machine would run with a dramatically higher luminosity than during “Run I” which ended in 1995. During Run I, about 140 inverse picobarns of collisions were generated, at the rate of about 3 inverse picobarns/week by the end of the run. The plan for Run II was for the new machine to get to a luminosity of about 17 inverse picobarns/week soon after the recomissioning, and ultimately to reach about 100 inverse picobarns/week.

This plan turned out to be wildly overoptimistic, as it took nearly a year and a half to get the machine operating even at the luminosity achieved during Run I. During FY02 the initial hope was to accumulate 320 inverse picobarns, rising to 830 in FY03 and 1300 in FY04. Instead 80 were produced in FY02 and 330 in FY03. While in FY02 the machine performed much worse than planned at the beginning of the year, in FY03 it did slightly better than planned.

As this fiscal year (FY04) is coming towards its end, the machine is doing significantly better than planned at the beginning of the year. The last few weeks have seen about 13 inverse picobarns/week being produced. Another measure, the luminosity at the beginning of a “store”, reached a record value on Monday of 8.28×10^31 cm^2/sec. This is about five times higher than achieved at the end of Run I. You can follow the progress (as well as the trials and tribulations) of the Fermilab accelerator physicists through on-line daily reports and continually updated luminosity charts. For a collection of documents showing the history of the Run II problems and the most recent estimates of what will be achieved, see the proceedings of the latest review of the Tevatron luminosity, which took place last February.

About 2000 inverse picobarns of collisions will probably be needed before the Tevatron experiments can push up the current experimental limit on the Higgs mass (114 Gev) that comes from experiments at LEP. There now seems to be a good chance the Tevatron will get to this point before the LHC starts operating at much higher energy in FY 2008. If the LHC achieves anything like its design luminosity, it will quickly make the Tevatron obsolete. Then again, having seen how hard it was to get the upgraded Tevatron running, the job of commissioning the LHC may not be so easy…

There’s a quite remarkable article by Barry Mazur in the latest issue of the Bulletin of the AMS. It brings together ideas about elliptic curves and deformations of Galois representations that were used by Wiles to prove Fermat’s last theorem, mirror symmetry, quantization, non-commutative geometry and much more. I’m not convinced it all hangs together, but it’s a wonderful piece of expository writing.

Mazur claims to be inspired by a very interesting seminar held every week in the Harvard math department called the Basic Notions Seminar, parts of which have recently been put online. This issue of the Bulletin is dedicated to the great French mathematician Rene Thom, who died nearly two years ago. The articles by Michael Atiyah and Dennis Sullivan about Thom’s work in topology are well worth reading.

At an early stage in the Los Alamos preprint archive it was split up into hep-th (for more formal or speculative work not directly relevant to experiment) and hep-ph (for “phenomenological” papers directly related to experiment). Susskind has just come out with his latest and now seems to feel that his ideas about the “Landscape” are directly of interest to experimenters and so belong in hep-ph.

The preprint is riddled with typos, for instance the third paragraph starts like this:

“During the last couple of years an entirely new paradigm has emerged from the ashes of a more traditional view of string theory. The basis of the new paradigm is the stupendous Landscape of sting [sic] theory vacua — especially the non-supersymmetric vacua. These vacua appear to be so numerous that the word Discrtuum [sic] is used to describe the spectrum of possible values of the cosmological constant…..”

You get the idea.

Some high points of the article:

1. “low energy supersymmetry – an ugly solution” to the naturalness problem. Now he tells us. From what I remember the “beauty of supersymmetry” has always been one argument made in its favor.

2. “the ashes of a more traditional view of string theory”. It seems that the picture of the world according to string theory that has been heavily sold for the last twenty years has burned down to the ground.

3. The argument in his last paper, such as it was, was wrong. Now he’s got a new one with a similar conclusion.

4. “… a prediction that supersymmetry will not be seen at the TEV scale seems warranted”. OK, string theory is finally making a prediction.

5. “If it turns out that low energy supersymmetry is a feature of TEV physics, then we will have to conclude that other considerations outweigh the counting of vacua on the Landscape”. So, even though string theory predicts no low energy supersymmetry, if it is found it doesn’t mean string theory is wrong. Got it?

Recently I’ve been reading a new book, The Evidence for the Top Quark, by a philosopher of science named Kent Staley. It’s a combination of a history of the CDF collaboration’s work leading up to their claim to have discovered the top quark, together with an extensive discussion of issues in the philosophy of science raised by the different methods used to analyze the data. The book is very topical since last week the D0 collaboration published an article in Nature claiming a new, more accurate, mass for the top quark based upon a re-analysis of their data from Run I of the Tevatron, which lasted from around 1992-96. The old analysis of the D0 data gave a top mass of 172.0 +/- 7.1 Gev, the new analysis gives 179.0 +/- 5.1 Gev. Combining the D0 data with the CDF data, the old analysis gave 174.3 +/- 5.1 Gev, the new 178.0 +/- 4.3 Gev.

Measuring the top quark mass is quite tricky since there are not a lot of events to work with and one needs to precisely measure the energies of jets. If a linear collider ever gets built, it would allow much more precise measurements. Knowing the top quark mass accurately is very important for the following reasons:

1. In the standard model one can try and use precision measurements of the electroweak parameters to observe the effects of higher loops including the Higgs and get a prediction for the mass of the Higgs. This crucially involves the top quark mass, since that is the strength of the top-Higgs coupling and the top quark couples far more strongly to the Higgs than any of the other fermions. With the latest D0 value for the top quark mass, one now expects (95% confidence level) that the mass of the Higgs is less than 237 Gev. For more details see websites at CERN and Fermilab.

2. In the minimal supersymmetric extension of the standard model there is an upper bound on the mass of the lightest neutral Higgs, a bound that depends strongly on the top mass. There’s an explanation of this on Jacques Distler’s weblog. With the newest value for the top quark mass one expects that the Higgs mass should be below 140 Gev in the supersymmetric case.

There are a few funny things about this report from D0:

1. It was published in Nature rather than the more conventional Physical Review Letters. Nature is not where high energy experiments normally announce their results and this appears to be an attempt to get wider publicity than is usual for such a result. For some comments on this, see David Harris’s weblog.

2. The new method of analysis is similar to one discussed extensively by Staley in his book: the “dynamical likelihood method” due to Kuni Kondo. Ten years ago the CDF collaboration was rather skeptical of this method and decided not to use it, basically seeing it as too complex to be reliable. Have they changed their minds? Will CDF re-analyze its Run I data using this technique too?

3. Much is made in the paper and the associated Fermilab press release that this result changes the “best estimate” of the Higgs mass from 96 Gev (which is excluded by LEP results) to 117 Gev, which isn’t. While this sounds impressive, it would be a lot less so if you do what you are taught in high-school physics and quote error bars with your numbers. As mentioned above, while it is true that the new result is that 117 Gev is “most likely”, it is also true that a very wide range of values is almost equally likely. A more sensible but much less impressive way of saying things would be to just say that at 95% confidence level the Higgs mass has to be between about 50 and about 250 Gev.

Update: The D0 Nature article is now at the arXiv.

According to a press release from UCSB, three theoretical physicists have proposed “the most viable test to date for determining whether string theory is on the right track”. This is based on a paper about cosmic strings where the authors manage to cook up a highly unlikely scenario where large strings exist and produce gravitational radiation observable by LIGO in the next couple of years.

Normally in the English language, calling something a “test” of a scientific theory would indicate that if it doesn’t work the theory is wrong. When LIGO doesn’t see this effect in the next two years I kind of doubt that there will be wholesale abandonment of string theory.

From one of the comments here I see that the Bogdanovs have put the reports on their theses on the CERN document server. One should perhaps take these with a grain of salt given their source. For instance, I wouldn’t be surprised if some reports were missing.

I’ve always had some sympathy for the people who ended up on the Bogdanov’s thesis committees. It’s a difficult position to be in when you have to decide what to do with students who seem to be enthusiastic and have worked hard, but are very weak and have completed not very good theses. A not unreasonable thing to do under the circumstances is to do one’s best to find something of value in their work, and leave the job of keeping nonsense out of the literature to journal referees.

But the Bogdanov theses, especially Igor’s, were so full of egregious nonsense, in particular with respect to topological quantum field theory, that they should have been beyond the pale. While some of these reviewers were string theorists, others weren’t, so the whole mess can’t be blamed on string theory.

Louis de Branges is a mathematician at Purdue who has had a long history of claiming proofs of the Riemann hypothesis. His latest claim has lead to a press release from Purdue. The press release points to what seems to be an older manuscript by de Branges outlining some of the history of the Riemann hypothesis and his work on it. This also includes some history of his ancestors, and de Branges has taken to calling himself “Louis de Branges de Bourcia”. He more or less promises that if he wins the Clay million dollar prize for solution of this problem he would use it to restore the ruined chateau de Bourcia for use as a mathematical research institute.

The actual purported proof is here. One mystifying thing about it is that in the abstract and introductory paragraph it repeatedly refers to relations to quantum mechanics, but there seems to be nothing about this in the body of the paper. Weyl’s book on quantum mechanics and group theory appears in the references, but nothing in the text seems to refer to this.

de Branges has a checkered history as a mathematician, with several of his claimed proofs of the Riemann hypothesis and other problems turning out to be incorrect. On the other hand, he did produce a correct proof of one well-known problem, the Bieberbach Conjecture. In that case his initial manuscript was pretty impenetrable, but after he explained his ideas to a group of Russian mathematicians, they gave a more understandable version of the proof and it became clear that de Branges really did have a proof. It looks like this one may also take some major effort to see what he really has.

For more about de Branges and the Riemann hypothesis, see the recent popular book “The Riemann Hypothesis: the Greatest Unsolved Problem of Mathematics” by Karl Sabbagh. A review of this book has some interesting comments about de Branges and his NSF funding.

A couple weeks ago a preprint appeared on the arXiv by R. A. Arenstorf, a mathematician at Vanderbilt University, claiming a proof of the twin prime conjecture. I asked one of my colleagues who is an expert on the subject about it and he said he didn’t believe it and would bet $100 it was wrong. Today I see that Arenstorf has withdrawn the preprint, saying that a serious error has been found.