I think the PRL issue, which had nothing much to do with the book being discussed here, has now been convincingly beaten to death. Enough about that.

During roughly the same time frame in which Wilson was doing his scaling-based work, Mandelbrot was formulating (and popularizing) fractals.

In my amateur opinion each was uninfluenced by, and possibly unaware of, the other’s progress. but in a sense they approached the same monolith from very different directions.

I take the point that Nature (or more generally CNS of Cell-Nature-Science) plays an important role in bioscience careers – but I don’t see that as a good example to be copied. I think PRL plays a similar role in condensed matter, but I agree with you this is not the case in high energy.

I have published in PRL (and on my own), but my best work is not there.

The thing I really don’t get though about emphasising PRL publications is that the PRL format is well suited to short papers making a single point. Maybe this works naturally for experimental subjects where there is a single clear result, and what matters is the measurement. But in theory there is an argument to be developed and calculations to be done, and in most cases 4 pages is not adequate to describe this. Such papers – and I think this represents most good work in theoretical physics – should not go to PRL.

My personal experience with papers within my conceptual radius of convergence is that what appears in PRL is disproportionately flashy, and more likely to be making big claims with a small chance of being right than making small claims with a large chance of being right (This is true of my own papers there as well). I don’t think there is any intrinsic sin in writing such papers, or that they are intrinsically better or worse than long calculation-heavy papers (which I also do), but to my mind this is the ‘PRL type’. So I see PRL more as a style of paper than an indicator of quality.

PRL emphasises general interest, which again affects the type of paper likely to be accepted, without to my mind necessarily implying higher quality.

I do think it is a good journal, and there is an approximate guarantee of quality, but I strongly feel it is suited to a particular type of paper that is not and should not be representative of most work in theoretical physics.

I agree with the comments about the ‘hotshot young theorist’ issue although I don’t think is especially to do with string theory (in recent years I would say this is much more prominent in BSM pheno). Particle theory as a whole, and particularly I think in the United States, has a problem with genetic diversity and inbreeding – look at faculty hires over a substantial period and see how many of these can be traced back to a very small number of advisors at Harvard/IAS. Yes these people are smart, but the history of physics shows how often truly brilliant people completely miss or misunderstand important new ideas, and it is not good if everyone is brought up with the same toolkit and prejudices.

I suppose Cormac is referring to No Go theorems that show that you can’t combine space-time and internal symmetries non-trivially, something that SUSY gives new options for. But this is getting way, way off topic, so please anyone who wants to discuss this with Cormac should contact him directly.

Your conversation with Piscador and others is also very interesting because I think the relation between theoretical physics and physics is at the heart of the debate around sting theory, yet this rarely comes to the fore. Peter touched on this quite a bit in his book, I imagine there’s quite a lot on it in yours.

As a very ordinary experimentalist who is the son of a well-known theoretician, I’ve long had a special interest in this debate. Sometimes I think theoretical physics occupies quite a unique position in science; because of its dual role in mathematics and physics (the former is often overlooked) it’s not quite clear whether it should be forced to conform with the strict positivism we tend to apply to other areas of science, i.e. should it always be shackled to experiment, or should some theoretical physics be allowed to roam free like mathematics. Perhaps the problem is when the latter is marketed as the former.

For example, SUSY is often portrayed on this blog as an offshoot of string theory. In fact, it was originally a purely mathematical construct that showed one way of cirmcumventing the no-go theorems of gauge theory – to this day, I gather than one cannot construct a unified gauge theory that includes the strong force without SUSY. It may have been adopted by string theorists, but that has nothing to do with the theory itself. You might of course argue that all this has nothing to do with experiment, but the original SUSY theorists never claimed it did – that’s my point!

As I understand it, the main motivation for SUSY nowadays comes not from string theory, but from cosmology, simply because it is very hard to see how the interactions were once unified without some sort of bose-fermion symmetry….
Regards, Cormac

“I have some sympathy for the points you make, but have never understood why when it comes to hiring the editors of PRL (typically non-practising physicists who are not subject specialists) and its referees (who are just a couple of members of the community) should play some special privileged role.”

Well, the editors of Nature are also non-practicing scientists and they have enormous influence on hiring in the bio-sciences…

How useful PRL is as a quality measure depends on how seriously the physics community is willing to take it. At the moment it does seem to be taken seriously by pretty much all of the community. String theorists profess not to, but I’ve seen firsthand how excited some of them get about sending a paper there when they think they have a shot. The quality of judgments on individual papers can fluctuate of course, but I think on a larger scale the level is pretty good.

As for its use in hiring decisions etc, I think it can play a useful role as a kind of minimum criterion. I’m sure that even some string theorists are sick of the phenomenon of the “brilliant young string theorist”, whose has been anointed as such by his illustrious advisor/mentor after being joint author on some papers with a bunch of more senior people. IMHO it would be a change for the better if such people first had to publish a few papers on their own in PRL before being anointed as `brilliant’ and sent to the front of the job queue. Just to show that they can. Shouldn’t take them more than a couple of months

PRL publications are also useful as an indicator of the overall health of a field. If interesting and important developments are happening in the field, it will manifest as a regular flow of PRLs. Absence of such a flow will be seen by the rest of the community as a sign of ill health. And if there is no PRL flow, and at the same time practitioners are seen to be actively hyping their field to the public, it looks really, really bad! And even worse when, on the rare occasion the practitioners do manage to get a PRL, they issue an excited and over-hyped press release…

FWIW I agree completely with what your wrote about the importance of ads/cft. On the other hand, I’m not sure about that 6-dimensional superconformal field theory that Nima is so excited about… If it’s really that interesting and important, he should go publish a paper on it in PRL!