On the Work of Edward Witten (1990) [pdf] | Hacker News Reader

On the Work of Edward Witten (1990) [pdf]

bohr.physics.berkeley.edu

39 pointsmontrose28 comments

Two years ago, I had the profound honor of having lunch with Prof Witten. He is also a very kind and generous human being. A small tidbit I learned: He spends a disproportionate amount of time thinking, and very little time on the desk with a pen and a paper. His students told me that he likes to work things out in his head -- which could last for days -- before he would commit things to paper. I thought that was interesting, although not unexpected for a genius at his level. If I am not mistaken, he is the most cited theoretical physicist.

He deserves to be celebrated in popular culture -- like Einstein and Newton. For me, he is the definition of a "living legend".

Physicist here. Just wanted to add a little context. Anyone in string theory will tell you that Witten is possibly the greatest physicist that has ever lived. Interestingly he's completely unknown to the general public.

> Anyone in string theory will tell you that Witten is possibly the greatest physicist that has ever lived.

Betcha experimentalists would disagree.

Not sure what you're talking about... The vast majority of experimentalists would not understand his work. That's why I used the qualifier "in string theory". You could as well have said that most biologists would disagree, would make about as much sense.

He probably means that theoretical physics is not the whole physics.

Sure, but these days it's where discoveries are made.

You know, in some sense experimental physics used to be "ahead" of theoretical physics. I accidentally put some current next to a compass and it moves. Why, no one knows.

I would say from 1905 onwards theoretical physics has been ahead of experimental physics. I have reasons to believe X, and a few years later experimentalists verify it. This difference is getting larger very fast. It took from 1933 when neutrinos were postulated by Fermi to being discovered in 1956. It took about 40 years from Higgs suspecting an extra particle was needed to it being found in experiments.

Maybe that's cheating a little but astronomy and astrophysics have had a good run of observations which physical cosmology then chases. Most of them after 1905.

High energy physics isn't the only physics. It's not even the only physics that uses the shiniest toys of theoretical physics.

Heh ... in grad school, one of my friends was a postdoc working on computing bounds for the Higgs mass. He used to tell me (theoretical condensed matter) that I worked in the dirt left over from the real physics.

Fast forward 20+ years, and he is working on modeling semiconductor growth (like I was to a degree), and I am long gone from physics. Skipped postdoc/tenure track to go into industry post Ph.D.

My impression dealing with HEP folks for many years, is that a fairly large fraction of folks in the field look down on the rest of the world. Similar with the nuclear people. This is anecdotal of course, and a small subjectively interpreted sampling.

All that said ... One of the more interesting aspects of string theory is that it appears to have analogues and applications in superfluidity, and other phenomenon where you get spontaneous symmetry breaking[1], or where you have critical exponents and length scales[2].

[1] https://arxiv.org/abs/1507.05635

[2] https://www.cambridge.org/core/books/string-theory-methods-f...

I beg to differ. You're right for the standard model. But there are other fields where this is not true. A few spring to mind:

- Solid state physics. There are entire fields which are not understood theoretically such as high temperature superconductivity

- Astronomy: dark matter and dark energy. There is no widely accepted theoretical model for both. Both where not predicted by any theory, either.

To say that theoretical physics has been ahead of experimental physics from 1905 onwards is not entirely accurate. In the 1960s particle accelerators had found lots of particles. At first those were thought to be elementary, but more and more were found, into the hundreds. Physicists started doubting that they were elementary. The quark model eventually explained that those hundreds of particles aren't elementary.

Saying that theory is ahead of experiment also suffers from survivorship bias. We tend to forget the theoretical predictions that didn't pan out. If the theoretical side predicts 10 different possibilities and the experiments eventually find that one of those is right, that's not evidence that theory is ahead of experiment. We also tend to interpret correct theoretical predictions in a more rosy light after the fact. One of the poster childs for theoretical predictions is the positron. It is said that Dirac predicted the positron, which is true. He first considered that an electron could be in a positive and negative state, but retracted that position later and said that the positive charges were holes in an infinite negative sea, and he theorised that the proton is a negative electron. He was eventually convinced by Oppenheimer that the infinite negative sea idea didn't make sense, and that's when he predicted that there was a positron particle in 1931. But the positron had already been observed two years earlier, although those observations weren't conclusive and the conclusive experiment was done in 1932. The interplay between theory and experiment is often nuanced.

What happens to that evaluation if string theory never has a successful experimental prediction? Is there some way his work would keep such a high standing as physics then? No argument that he must be incredibly smart, but that's not the same.

My layperson's impression is that string theory implies supersymmetry and supersymmetry doesn't easily comport with the latest accelerator results.

> What happens to that evaluation if string theory never has a successful experimental prediction?

Exact Lorentz is a very successful experimental prediction of string theory.

> My layperson's impression is that string theory implies supersymmetry and supersymmetry doesn't easily comport with the latest accelerator results.

Symmetry breaking.

That special relativity continues to hold at high energy is not a new prediction. It does distinguish string theory from some other guesses at new physics.

https://en.wikipedia.org/wiki/Supersymmetry_breaking_scale implies LHC was expected to show evidence of supersymmetry. That's what I was thinking of by 'easily'.

Anyway, I guess you are saying he's the greatest ever because of string theory? And we'll just have to see if that pans out as physics.

> That special relativity continues to hold at high energy is not a new prediction. It does distinguish string theory from some other guesses at new physics.

Moving goal posts. First you asked for predictions, now you're asking for "new" predictions.

Anyway, what does "new" even mean? Some theories say it holds, some theories says it doesn't. String theory says it does, and it's correct.

Well, sure, but the standard model also says it does. If string theory plus whatever additions are needed to make it "save the phenomena" (predict the observed world in quantitative detail -- which hasn't happened so far) turns out to be simpler than the standard model, then at that point it'd make sense to move from talking about new predictions. Otherwise we're talking about predictions that aren't really predictions at all, in the sense of separating string theory from competing theories such as the standard model.

I think I'm using a well-accepted understanding of science which I'd taken for granted, instead of moving any goalposts. A theory retrodicts some of the same phenomena as its parent theory, as an assumption: that's not evidence that the new theory is more true than the old theory. What I started the thread with was an operationalization of "what if string theory is not true?" It's not a perfect one: string theory could also become accepted as true by finding a way to make it retrodict only things already observed, but with a less complex theory than the standard model. But come on, we're very far from there.

In what sense does string theory predict any aspect of Lorentzian symmetry?

Vague memory from a non-physicist: there are some different proposals about new physics (e.g. quantum gravity), which violate that symmetry at high energies, which would predict dispersion of arrival times from far-away gamma-ray bursts or something, which wasn't observed. So string theory did better than those proposals, but didn't predict anything new here, as I understand it.

Witten's work has taught us an enormous amount about the structure of ordinary quantum field theory. It's important even without strings. Indeed, Witten was already a dominant influence in the physics community when he began working on string theory.

Thanks -- that's more the sort of info I was hoping to find out.

Another nugget: one of the leading experimental approaches to search for dark matter “direct detection” (which would be a Nobel prize worthy discovery) has its origins in a paper coauthored by Ed Witten: http://inspirehep.net/record/207030

Even if they're not in string theory, they'll probably say something like that, or perhaps imply that he's actually an alien super-intelligence...

Isn't there actually an xkcd for that? Can't find it now, maybe I'm imagining.

Greatest is a complicated word. But smartest is a real possibility. I think von Neumann was the last person who matched Witten's raw intellect.

My memory isn't as good as it used to be. I thought the article was about Edward Frenkel at first who wrote the fantastic novel Love and Math [1]. It's about his life growing up in Soviet Russia as a Jewish mathematician. He was good friends with the famed physicist Murray Gell-Mann who discovered a theory of hadrons, leading to the quark model we have today [2]. In the novel he talks about some of the groundbreaking work he has done. But he's mostly known for his cinema work on the emotional fullness of mathematics and physics. I would nominate him for a Nobel if I could. But Fields medals are about pure mathematical achievement, right? So I'm not sure he would qualify.

[1] http://amzn.to/2HKD3Mc [2] https://en.wikipedia.org/wiki/Eightfold_Way_(physics)

Frenkel was tenured at Berkeley in his early 20s, at an age when most people are still in grad school. His mathematical achievements are substantial. I've no doubt that he was on the list of candidates for the Fields at some point in the late 90s & early 2000s.

Small correction to the submission title: it's "Witten" not "Whitten"

Whoops! Thanks, fixed.

Reply on news.ycombinator.com