Every so often I’ve taken a look at something about theoretical physics on Reddit, generally ending up not spending much time there. One reason was that I realize I’m already spending more of my life than is healthy arguing with people about string theory and the like, so better to avoid a new venue for that. The temptation to respond is strong when one sees someone mischaracterizing one’s opinions, but I’ve generally been able to resist temptation at that site.
Today I happened to come across a really wonderful discussion there though, and wanted to draw attention to it, even though it’s from a year ago. It’s entitled A View from an Ex-String Theorist and consists of a long piece by someone who has recently left string theory, as well as some answers to questions asked by others. If you want to understand what string theory looks like these days to good theorists who are working on it, read what “No_More_Strings” has to say.
No_More_Strings explains very well the difficult job situation in the field, and the effects this has. With a lot of very smart people and almost no jobs, postdocs are in no position to take the time to try and learn something new that isn’t a “hot topic”, or try and work on an unpopular idea that might take years to go anywhere. This is a huge part of the story of why this field is in trouble, and the situation seems to have just gotten worse since I wrote about it in my book over 10 years ago.
The suggestion that “string theorists” should stop calling what they do “string theory” is an excellent one. No_More_Strings explains how smart people in the field are not working on string unification, but have moved on to different things with little relation to quantized strings. Giving up the name would be a good first step to allowing people to think of what they are doing in a less narrow way. If you didn’t have to start every grant application by explaining that you’re motivated by “our best hope for a theory of everything”, you might find it easier to work on something quite different, with no relation at all to quantized strings.
I can’t quite resist correcting a couple things mentioning me. No, I don’t think string theorists are stupid. No, I don’t think that Witten “singlehandedly destroyed the study of “real” physics” (the last one isn’t the fault of No_More_Strings).
If they are not stupid, then why on earth they are insisting on a program that they know (better than anyone else!) has no promising future?!,
You don’t seem to have read or paid any attention to the piece, which explains pretty well that what most “string theorists” are doing (at least the smart ones…) is working on a wide range of topics, of varying degrees of promise for various purposes, most of which have little to do with strings or the supposed use of strings to get a unified theory.
And, even those who haven’t given up on string unification are staying with it not out of stupidity, but out of an honest belief that it’s the best idea about unification that they know of, no matter what its problems. Anyone who thinks Witten is stupid is a fool…
I’m sure we’ll hear Lubos’ long reply about this tomorrow. Thanks a lot, Peter.
Gia Dvali has an interesting new understanding of black holes as condensates of gravitons with energy 1/R. If true, this would also mean that black holes are insensitive to UV physics (Einstein general relativity is enough); so computations of black hole entropy in terms of string micro-states would just be way a misunderstanding black holes.
I am consistently staggered by some of the insinuations that surface in rants like the one on reddit that you linked to of failed researchers (be it that they didn’t land a permanent job or that their research program collapsed).
.) Why do they always need the collective shoulder-patting of “we are the smartest”?
.) Why do they believe that their small subfield is so very special in terms of emotional, intellectual and time commitment?
.) Why do they believe that their particular area is starved of academic positions?
.) Why do they think they are too “smart” for the “dumb” homework problems or laundry tasks?
.) And finally, why do they assume that reading fiction loosely related to the theme of research qualify you for anything but a job as librarian?
I am always reminded of the students that are “bored by classical physics” and the like and jump into quantum field theory in year 2 or so. Never saw a single one of them being successful.
It’s a simple truth that research as a career is highly risky, extremely time consuming, very competitive and very often frustrating, quite independent of the specialization. I would have thought that – especially in good institutions – you get this message very early on and consistently.
It’s funny that he repeats how smart he is while complaining that all he has ever done is `laundry’. The academic system is designed to select precisely those people who are the best at `laundry’ and this guy seems to have indeed been very good at it. But there are many other necessities, besides technical skill, to be a really good scientist like creativity, courage, perseverance and luck.
The canonical example being Einstein, but even the originators of string theory had to toil in obscurity for a decade or so. If you don’t have the courage to just do what you’re interested in because you are addicted to the praise of your supervisor, you are afraid of losing your position or god forbid, not getting hired at a Top Ten university, or you don’t have any ideas that you truly believe in and can make progress on then you are lacking the tools.
It’s a shame that the system doesn’t allow everyone to just study whatever they want to for as long as they need to (though would that really lead to more insights or just more time for bad ideas?) but as it is this guy had found his level: good technically, not a lot of insight. I have never met a tenured professor who, if you had a really good idea would stop you from working on it, even if it slowed down your output. University is supposed to teach you technical skills that are necessary, but curiosity and creativity are going to make you a great scientist who is doing interesting things. If you don’t have those, and are also unwilling to take risks, then working on obscure technical problems is probably the best place for you.
Sorry, there is no reason to criticise the guy in question. Leave him alone! Yes, you have to take risks to do good research. But you also have to feed your body and later your family. The story of the young man is still happier than that of Lubos Motl, who also left string theory, and now earns his bread writing rants on his blog, for his own fan club. And probably the story is much happier than those of Seiberg or Arkani Hamed, who lure people into the field telling falsities. Everybody takes his choices in his life. Also Motl’s decision is a choice. Even the IAS people luring others to join them have taken a choice. Are they happy? They have to decide.
IMHO, we reduce the talk about human choices, but instead talk about unification, and on how to achieve it.
I am not criticizing his decision to leave physics, rather the unsaid assumption that because he is `really smart’ boring technical work is beneath him. It’s a shame about the job situation but doing the exciting, creative research that he seems to want to do instead of `laundry’ would have landed him in exactly the same position. So he didn’t have either the guts or the ability.
I think you’re being too hard on “No_More_Strings”. I didn’t read his (seems highly likely it’s a “he”….) piece as complaining at all. It provides a very accurate take on what the grad student/postdoc system is like in that particular academic field. People who choose to enter it likely have a fairly good idea what they’re getting involved in when they start, and reasonably good career prospects for doing something else when they leave (and this is what will happen to almost all of them). More interesting is the significance for the field itself: is this organization of research, this particular set of incentives and pressures, a good way to encourage progress or not?
Maybe I am being too harsh but his attitude about `laundry’ problems and his repeated assertions about how smart he is are galling. If you don’t like it you either quit, as he wisely did, or take a risk and do something interesting. Most people (including me) don’t have what it takes to succeed at the latter.
Regarding the organization of research, have a look at http://www.nature.com/news/the-future-of-the-postdoc-1.17253 . From personal experience I have seen an institute hire half a dozen permanent staff and even more postdocs without increasing the quality of its output. The ‘publish or perish’ system is often maligned but what is the alternative? Don’t publish but keep your job forever? The truth is that boring technical laundry is all that most people can do and the various (also maligned) metrics like citation tracking and impact factors do a reasonable job of measuring quality.
If there is a problem it is minting too many PhDs like this guy, with technical skill but an inability to drive research themselves who are unlikely to become good scientists. The blame for that is principally on PIs and university administrators who make hiring PhDs and postdocs appealingly cheap. Progress will happen and is happening, it is just harder to distinguish from background noise and perhaps a few bright people are lost to the banking industry.
It’s fascinating as a mathematician to see this discussion. Math and physics are close, but the job markets have always been very different, since math is a service department and hence universities need a lot more mathematicians. It’s not that the job market doesn’t suck, it does, but nowhere near as badly as in physics. If you go back to when I was in grad school (late 80s), everyone got an academic job when they got their doctorate. Most of them weren’t serious research jobs of course, but you could do some research in most of them. And almost all the jobs were tenure track. It started falling off a cliff the year I got my Ph.D. (great timing on my part), but even so, most of the people I know got jobs, tenure track, just later than they would have, and not in places as good as they were hoping for. My thesis was published in a very very good journal, and when it got seen I actually got asked to apply to a top program – didn’t get it, but I did have a shot. If I hadn’t have been married, and settled in NYC, and if I could have moved anywhere, I might well have been able to move up to a pretty serious research program. As it is, I have tenure, I enjoy my job, I’ve done research, and it’s been supported. From what I know this is extremely unlike physics.
Part of the problem is pretty obviously physics should produce many fewer Ph.D.s They’re just setting people up for 5 or 10 years of stressful postdocs with almost no chance of a permanent job, and very little opportunity to work on what interests them. Most of them end up doing the laundry, and eventually hating it. An alternative I guess would be not to drop the number of doctorates, but instead just let postdocs do whatever they want. Most still wouldn’t get jobs, but they would enjoy what they were doing, and maybe it would drive research in interesting directions.
One other interesting thing is I really think there is much less laundry in math. Not sure why, maybe it all got done in the 17th Century. No one wants postdocs to sit and do boring stuff for them, it wouldn’t help the senior people or the junior people.
I also found the contrasts to math quite interesting, including the comments about this and the comments from a mathematician there.
One thing that struck me is that my experience as a grad student/post doc didn’t at all involve a senior person making me work all the time on endless computations. My main memory of the time was of how difficult it was to get a senior person to take any interest at all in what I was doing. I suppose this is just evidence I wasn’t doing it right: unless you’re an obvious genius, the way to the small chance of a successful career there may very well be finding a senior person who needs this kind of work done, under time pressure, and get it done.
I also don’t see that happening very much in math, where the pressure on young people seems to be more to come up with something oneself, to show that one can do math research oneself at a high level. Part of what is different is the very different job situation.
I often advise young people trying to decide which field to go into that at this point they’re more likely to be able to have a career doing work related to deep issues in physics in a math department than in a physics department, and this piece provides more evidence of that. The author’s attitude of “I don’t do proofs, I do do laundry” I don’t really understand.
You’re right about the fact that for young folk in math the thing to do is to show you can do something yourself. Even your thesis. And you never know what it will be. The thing in my thesis that actually got attention was a technical lemma, not the actual result.
The first commenter is spot-on though – “I think the fact that you never really enjoyed learning quantum, E&M, etc was a big warning sign.” This person says they didn’t much enjoy most of physics, so why am I surprised that their career in physics didn’t work out.
When I am admitting people to grad school a really useful “rule of thumb” is whether the person is more excited about the idea of being a physicist than with physics itself — since that is a red flag.
Pingback: Confessions of an ex-string theorist | Uncommon Descent
I was interested in your comments about advising students. If you’re a young student about to enter graduate school with interests in mathematics and physics, and you’re having a tough time picking between the two, do you have any advice on how to choose?
If you’re interested in an academic career after a Ph.D., and find math and physics equally appealing, go to graduate school in math. You’re just much more likely to have a chance at a reasonable academic job, and math is intellectually a very healthy subject. If you’re interested in the experimental side of physics, that’s very different than what mathematicians do. The only way to pursue that is in a physics (or maybe engineering) department. If you’re not set on an academic career, but interested in getting a Ph.D and then doing something other than teaching at the college level, then there’s no particular reason to necessarily favor math over physics.
I may add that there is also condensed matter physics, which is a quite healthy subject. There are many interesting problems in theoretical and computational condensed matter physics (and in theoretical chemistry). These problems have usually straightforward connections with experiments. A PhD in these fields can be quite satisfactory for students with interest in the theory side. The job market does not seem terrible, at least in Europe, so that academic or research career is a difficult but still reasonable option.
P.S. Peter, your blog is a fantastic source of information. I have been following it for several years but it is the first time that I post a comment.
You didn’t ask me, but people often ask me this, so my page of advice to the young scientist includes a section “Math or Physics?”
This is not about the job situation: it’s about the big differences in research in these two fields, which may not be apparent to students who like courses in both subjects.
I had not read your “advice …” Good stuff. I wonder where “No More Strings”‘s “raw naive curiosity” went. Maybe his/her lack of much of this is why he/she decided to get out. What you say about talks is pretty apposite, and the one thing that ought to have persuaded me that I had no future in physics was that I thought 95% of seminars were a waste of time. The most common fault was simply that they assumed that you, too, had been working on a similar problem for the last 6 months. After the first slide nothing much was comprehensible, but no-one in the audience, including me would dare admit that for fear of being thought ignorant. God knows how many hours of my life (and other audience members) went down the toilet as a result. Personally, I would rather outrage the audience than bore them. At least they will remember it.
For me, one of the more sobering aspects of this reddit story comes in the comments, where reddit user Valeen discusses his move from string theory to the oil and gas industry. I think there’s a big problem if the pipeline of talented young string theorists is being diverted into figuring out ways to burn _more_ of the fossil fuels in the earth.
condensed matter and John,
Thanks for the help with advice to Justin. I should have noted that I’m woefully uninformed about the job situation in condensed matter theory, even more when it comes to Europe or other non-US areas.
You might want to keep in mind that probably a lot more ex-HEP theorists end up working in the financial industry and for the NSA or other “Homeland Security” businesses than work for energy companies…
Doesn’t the existence of the math department hinge more on the intro calculus sequence, remedial courses, differential equations and linear algebra though? Their actual research output is generally of little interest to the outside world short of what goes on in the applied math department. The abundance of math positions then would seem to be furnished by the fact that you can teach intro calculus and ponder the Goldbach conjecture without needing to find a grant to fund your lab equipment or personnel.
In other words, mathematics only possesses superior job opportunities if you are comfortable with being irrelevant; otherwise you’re an applied mathematician competing for supercomputing grants like everyone else.
Sure, the job situation in math is reasonable because of the needed teaching (if you despise the idea of teaching students math at that level, then you should try another field). Not being dependent on grant money has definite advantages. Even on the applied side of math though, while grants are important, there’s nowhere near the same imbalance between the number of talented people and the number of jobs as in areas like string theory/HEP theory.
I had in mind contrasting HEP theory vs. mathematics, and was commenting on the job situation. If you instead want to debate who is more irrelevant and unconnected to the real world, it’s not at all clear to me that these days mathematics can win that one…
Haha fair enough, HEP and much of pure mathematics seem to have considerable trouble relating to the mundane.
I wonder if part of the problem with the modern HEP research paradigm is that it attracts individuals with an attachment to exotic physics/mathematics. Not that such things may not be necessary, thus far no one has formulated GR or QFT without the exotic physics/mathematics, but rather that one seems to get a lot of individuals with an emotional attachment to such things, and if there is a better way to interfere with a scientist’s good judgment than to let him/her develop an emotional attachment to an idea I cannot think of it.
Most of the students at my not-really-prestigious university who want to be string theorists are swayed by the perceived glamour of the field. It’s quite clear talking to them that they are not interested in actually doing physics–they want to come up with that Next Big Idea. Not coincidentally, these are usually not very talented students either. The good students (in our population–the situation is probably different at “better” universities) are more generally interested in challenging problems no matter the subfield. They usually go on to get some sort of good job, if not in academia than at a national lab. (My latest student, who is defending today, has competitive offers from 4 labs.) My own field, nuclear physics, is one that many string theorist may sneer at, but I have steady funding, I work on problems I find interesting and satisfying, I have fun and interesting colleagues from around the world, I get to go to conferences around the world. And the Department of Energy actively encourages universities to hire in my field, as support for their experimental facilities. The competition for jobs is still tough, of course–the grass is no more greener here than elsewhere. True, it’s not as glamorous as string theory or particle physics in general–but if and when we actually detect non-baryonic dark matter, or lepton nonconservation in neutrinoless double-beta decay, or many other experimental tests for beyond the standard model, you’re going to need nuclear physics, and good, reliable nuclear physics, to make sense of those experiments. Maybe that’s “laundry” physics, but it is still fun and satisfying.
Sadly I must admit that I used to be just such a student and was quite caught up in the massive hype surrounding HEP. I think part of the problem is the publicity campaign; I was interested originally excited about more “mundane topics”, especially polymer physics and biology, but when I heard individuals like Brian Greene wax lyrical about the grandeur of theoretical HEP I had a hard time justifying an interest in something as comparatively “unimportant” as protein folding or percolation. Luckily for me I accidentally stumbled back into my interests and never looked back; a pity, I think, that others are not so fortunate.
Pingback: A View from an Ex-String Theorist | Dr. Goulu |...
The picture I get from your comments is not really encouraging for someone who wishes to contribute to theoretical physics.
If theoretical physics seems to cost a lot of money and yet produce very little these days, why not just stop paying people to do research and pay them only for teaching, just as it was done in the past centuries. Good ideas will then certainly come form here and there every now and then. But, since students are supposed to write something before being awarded a degree, why not just let them investigate those ideas: if they don’t find anything useful about them, they just report it and they get there degree. If they find something useful, they describe it, they get their degree, and they go on teaching with their diploma and continue working on the idea on their own without getting payed for it. This way, nobody will get payed for doing research.
Is this way of thinking too naive?
I also wonder if the same grey picture is seen elsewhere in some sub-field of other sciences.
Why do you say theoretical physics produces very little? There is tremendous vibrancy in theoretical physics, especially if you don’t make the mistake of thinking theoretical physics = HEP (or worse yet, = string theory only*). But there is wonderful work being done in nuclear physics, astrophysics, cosmology, biophysics and polymer physics, condensed matter, complex systems…and in HEP, just not only HEP. The job market is tight, but it has been tight since the 1970s or before.
*PS — I am not against string theory, I am against “string theory and nothing else.”
Sorry, by theoretical physics I meant fundamental physics.
So, in other words, you make the mistaken equality. Okay, very well. I don’t agree, but let’s go with it. Are you suggesting that because “fundamental theoretical physics” is in a bit of lull, that all students of theoretical physics, including “non-fundamental” subfields of theoretical physics that are in fact quite productive shouldn’t be paid either?
Also, when you say that the endeavor in “fundamental physics” (I refuse to accept the equality with theoretical physics) costs a lot of money, are you talking about just theory or experiment? As far as theory, only a tiny amount is spent on it. A few years ago the NSF budget for high-energy theory was about $1 million / year, at least for ordinary grants, with about half going to string theory. Maybe about the same from the Department of Energy. But a lot more money goes into experimental physics, and certainly a lot more goes into even other fields of theory such as condensed matter theory. So, relatively few students get supported to pursue theoretical “fundamental physics,” and many more get supported to pursue other subfields of theory and even more to do experiment.* So, I think your scenario _is_ naive, or at least not really reflective of how physics is pursued and funded in the US.
*For what it’s worth, even in these fields money is very tight and very competitive and I’m not arguing there are just buckets of money lying around. But anyone who thinks string theory is sucking up all the money in the room and failing to produce a final theory for that, is mistaken.
“Are you suggesting that because “fundamental theoretical physics” is in a bit of lull, that all students of theoretical physics, including “non-fundamental” subfields of theoretical physics that are in fact quite productive shouldn’t be paid either?”
No, he means fundamental theoretical physics only. I thought that was clear from his most recent comment. He’s making no statements about non-fundamental theoretical physics.
Theory is relatively cheap compared to experimental physics (although I think the HEP theory NSF budget is more like $10-15 million/year, DOE maybe $50 million/year, compare to HEP experiment of maybe $800 million/year). Much of what this pays for is postdocs, the real problem is the small number of permanent jobs compared to the number of people going into the field. Most permanent jobs are funded not so much by grants, but by university budgets, so funded now out of tuition revenue.
Given university budgets, there’s no way you’re going to see a big expansion in full-time reasonably paid permanent teaching positions. In principle physics department could decide to hire more in HEP theory, less in other fields, but the trend is the opposite.
I agree budgets have been static for a long while so no chance of big expansions. DOE likes to fund theory relevant to their experiments, which nowadays means RHIC, J-lab, and the under-construction FRIB and others. They even offer carrots in the form of “bridge funding” to encourage universities to hire faculty in a certain area. But even that amount is relatively small, and it corresponds only to a mild swelling distributed over many universities, and the competition for these few additional spots–one, maybe two new ones in a very good year–is very tough.
Dear CWJ, Vlad, and Peter,
I thank you very much for all your comments. Vlad had exactly my point.
Indeed, all your comments give me a more accurate picture of how physics is pursued and funded in the US.
I don’t think today’s physics job market environment is that different from when I was a undergraduate physics major in the early 1970’s with dreams of eventually making a big discovery in theoretical particle physics. (I was at MIT during the charm discovery of the J/Psi – November 1974 Revolution no less!). Fortunately or unfortunately, my travails in the required MIT undergraduate quantum mechanics course convinced me to move out of physics into electrical engineering, and then on to finance.
Over the last 40 years, I have always heard and read about the employment problems of newly graduating PhD physicists, even as I thought about what could have been, as my physics interests continued from afar.
there’s also a lot of fundamental physics to be explored outside of high energy physics. All sorts of collective quantum phenomena that emerge within interacting many-body systems are no less fundamental than, say, quantum gravity. Or quantum information, which is even more fundamental than physics itself! The techniques people use in these fields are often quite similar to those used in HEP, with the main difference being that there are actual experiments that you can compare your results to.