The Future of "Lone Physicists" - Researching Physics Alone?

[Loststudent22]

Slightly off topic but why is massive collaboration so common in the sciences but not in mathematics? For example in physics the higgs boson discovery had like 3k authors, where in math its almost always been 1 or 2 authors working in isolation. For example Andrew wiles working for 7 years in secret on Fermats last theorem before collaborated to fix the proof with richard taylor or perelman working alone for years to prove the poincare conjecture and even yitang zang working in secret to prove the gap between primes. Those are some of the biggest recent discoveries in math also.

 

[Shablong]

Slightly off topic but why is massive collaboration so common in the sciences but not in mathematics? For example in physics the higgs boson discovery had like 3k authors, where in math its almost always been 1 or 2 authors working in isolation. For example Andrew wiles working for 7 years in secret on Fermats last theorem before collaborated to fix the proof with richard taylor or perelman working alone for years to prove the poincare conjecture and even yitang zang working in secret to prove the gap between primes. Those are some of the biggest recent discoveries in math also.

It's just the nature of the work. Mathematician's are playing with their beautifully ideal universe. The real world of physics, however, has many more surprises.

Edit: Also... as for your reference to "..3k authors." That is due to very large experimental collaborations as in CERN. Mathematicians don't need billions of dollars of equipment and thousands of people to maintain and use it. Most of the time, as I have found, theory papers tend to have less people to which credit needs to be given. Experiments definitely require collaboration.

 

[jtbell]

Experimental physics papers usually list as authors all the professors and graduate students who were involved. Experimental particle physics is done by collaborations among groups of professors and grad students from many institutions. All of them get listed as authors on the papers that the collaborations produce.

During my time as a graduate student in experimental particle physics about 35 years ago, it was customary in that field for all authors to be listed in alphabetical order. When I joined my Ph.D. research group and became part of two collaborations with about 35-40 people each, my name started to appear first on all papers published by one group, because I happened to be first in line alphabetically.

 

[Niflheim]

Just out of curiosity, for theoretical papers how is it determined who is listed as an author? Obviously the main developer(s) are listed, but what about people who just help a little bit or who are in the group but don't actually work on that paper, like if the group is large?

 

[Jano L.]

That is what I assumed you meant.

It doesn't happen now, and I am not sure that it ever happened. Einstein certainly did not work in isolation, and even the single author works of today still come from people working on theory with other people. Perhaps Newton, I am not sure.

Of course it has happened with Einstein. Working at a patent office and occasionally discussing physics with friends in his free time is not an isolation, but it is not a group research project either. He was working by himself. It surely is going on nowadays too, the people who do it are just not very famous, because they do not value publishing and publicity as the research project oriented people do.

 

[Choppy]

Just out of curiosity, for theoretical papers how is it determined who is listed as an author? Obviously the main developer(s) are listed, but what about people who just help a little bit or who are in the group but don't actually work on that paper, like if the group is large?

It's field specific, but usually, the first author is the person who made the largest contribution to the work, usually the one who did the bulk of the writing, and often the person who has agreed to act as the corresponding author (the person readers should contact first for follow-up inquiries). Sometimes the last author is reserved for the "senior" person or the person who facilitated the project. But that's not always the case. Sometimes author order is a toss-up, and sometimes it can result in heated debates.

It's important that the people who get their names on the paper have made a significant contribution to the project - this usually means going beyond providing some data or putting together a graph (and technically it also means more than securing the funding for the project). What I tell my students is that by becoming an author you are agreeing that you are responsible for what's been published. That means that anyone should be able to come up to you at a conference and ask you about the details of the work.

Unfortunately though that PhD comic I posted earlier was a tongue-in-cheek comment, there is a kernel of truth to it.

 

[PeterDonis]

when I say lone physicist I mean someone working on the actual theory by themselves, ie not in a group research project.

Judging by the continued disagreement about who qualifies, I'm not sure how much this really clarified the definition.

It might help if you would say whether some of the specific examples being discussed in-thread meet your definition of a "lone physicist". For example, Einstein working out special relativity while working as a patent office clerk. (Btw, I'm not sure that Jano L's description, "occasionally discussing physics with friends in his free time", fully captures Einstein's use of his friend Besso to discuss his ideas with as he developed SR.)

 

[Choppy]

I suspect part of the motivation behind the initial question may come from experience with group work - where group of students are assigned a project and the student is question is forced to deal with the dilemma of picking of the slack from others who are not carrying their weight or risking a low grade. While such experiences can serve as opportunities for developing social and co-operative skills, they can in some cases lead to less-than-memorable experience and leave a bad taste in one's mouth for careers that involve group work.

But it's important to remember that the further you go, the better at collaboration people get. You develop a specific skill set and eventually people seek you out for that skill set. You move from random assignment-type collaborations to collaborations that you choose to be a part of, and if someone in the group is not pulling his or her weight, they don't get asked back.

 

[Niflheim]

I suspect part of the motivation behind the initial question may come from experience with group work - where group of students are assigned a project and the student is question is forced to deal with the dilemma of picking of the slack from others who are not carrying their weight or risking a low grade. While such experiences can serve as opportunities for developing social and co-operative skills, they can in some cases lead to less-than-memorable experience and leave a bad taste in one's mouth for careers that involve group work.

But it's important to remember that the further you go, the better at collaboration people get. You develop a specific skill set and eventually people seek you out for that skill set. You move from random assignment-type collaborations to collaborations that you choose to be a part of, and if someone in the group is not pulling his or her weight, they don't get asked back.

That's about right actually. Part of it is curiosity and part of is the exact reason you stated. I haven't had the best experiences with serious group work thus far in my education, and I can so easily see myself in an undergrad or early grad research group doing a paper almost entirely by myself with a bunch of others freeloading and me having to include them as an author. I understand what you're saying, and agree with it, and know it will be the case in almost all circumstances, but I guess I just needed to confirm it.

 

[mfb]

It's important that the people who get their names on the paper have made a significant contribution to the project - this usually means going beyond providing some data or putting together a graph (and technically it also means more than securing the funding for the project). What I tell my students is that by becoming an author you are agreeing that you are responsible for what's been published. That means that anyone should be able to come up to you at a conference and ask you about the details of the work.

Experimental particle physics gave up trying that. There is just no fair way to tell who made a significant contribution to a specific paper, so the whole collaboration gets listed (over 1000 authors for ATLAS and CMS). At more than 100 papers per year, most members won't even recognize all the titles of the papers they are listed on.

Independent of author lists: talking to other physicists is an important part of research, both on the theoretical and the experimental side.

 

[Dale]

Of course it has happened with Einstein. Working at a patent office and occasionally discussing physics with friends in his free time is not an isolation, but it is not a group research project either. He was working by himself. It surely is going on nowadays too, the people who do it are just not very famous, because they do not value publishing and publicity as the research project oriented people do.

I don't know. He wasn't a professional physicist at the time of the development of SR, but I don't think that means it was done "solo" either. Certainly, by the time he developed GR, he was working in a team. I don't think that he is a clear example of a solo physicist.

EDIT: see russ waters' comments below. I wouldn't count him as "solo" even for the SR portion of his career.

 

[Dale]

I haven't had the best experiences with serious group work thus far in my education,

That is a problem. Teamwork is critical for all but the most trivial tasks and all but the most menial jobs. Nobody wants to hire people that will not function well on a team. That will seriously limit your employability, as well as harming your own personal happiness.

 

[russ_watters]

Of course it has happened with Einstein. Working at a patent office and occasionally discussing physics with friends in his free time is not an isolation, but it is not a group research project either. He was working by himself.

Einstein was a physics phd candidate when he developed SR (I don't know why people always leave that out and focus on his work at the patent office - to me, it's more relevant than how he made his money). SR wasn't his thesis, but his connection to the physics community was a lot tighter than "occasionally discussing physics with friends."

 

[sanjid]

there aren't such rules actually. if you want to encounter your physics by own self, its fine. exploring physics is always a fascinating stuff. but nowadays, if you want to figure something out very special or fundamental you need to work jointly. physics has gone so far. tiny things have been found already. now there are bigger insights concealing for your concentration. so if you work, share your ideas and unify some other's related work; success will come soon :-)

 

[Niflheim]

That is a problem. Teamwork is critical for all but the most trivial tasks and all but the most menial jobs. Nobody wants to hire people that will not function well on a team. That will seriously limit your employability, as well as harming your own personal happiness.

I mean with the group itself, I'm generally good working with others.

 

[Dale]

I mean with the group itself, I'm generally good working with others.

Oh good. Then I misunderstood.

In that case, it is not a question of being a solo physicist, but rather just a question of finding a good group. That is always a concern, but chances are that you can make good contacts at conferences and get a feel for how the group is.

 

[e.bar.goum]

Oh good. Then I misunderstood.

In that case, it is not a question of being a solo physicist, but rather just a question of finding a good group. That is always a concern, but chances are that you can make good contacts at conferences and get a feel for how the group is.

In addition, I think that the quality of groups in academia are better than in undergrad/high school (on average). Academia selects pretty strongly for this.

 

[atyy]

Slightly off topic but why is massive collaboration so common in the sciences but not in mathematics? For example in physics the higgs boson discovery had like 3k authors, where in math its almost always been 1 or 2 authors working in isolation. For example Andrew wiles working for 7 years in secret on Fermats last theorem before collaborated to fix the proof with richard taylor or perelman working alone for years to prove the poincare conjecture and even yitang zang working in secret to prove the gap between primes. Those are some of the biggest recent discoveries in math also.

Is what you are saying even true? For the sake of discussion, let's take authorship as a practical measure of "working alone". The Atiyah-Singer theorem has already been pointed out as a major mathematical discovery obtained by a collaboration. Another that comes to mind is the Green-Tao theorem. Below are descriptions of the work of last year's Fields Medalists. It looks like Hairer was the only one who worked alone. Incidentally his prize is the most physiky in some sense, or maybe the least - I can't understand his work at all - I'd love to understand it, because apparently he is able to give a proper mathematical existence to an equation that condensed matter physicists have been studying for years.

http://www.mathunion.org/fileadmin/IMU/Prizes/2014/news_release_bhargava.pdf
http://www.mathunion.org/fileadmin/IMU/Prizes/2014/news_release_avila.pdf
http://www.mathunion.org/fileadmin/IMU/Prizes/2014/news_release_hairer.pdf
http://www.mathunion.org/fileadmin/IMU/Prizes/2014/news_release_mirzakhani.pdf

On the other hand, in physics the following important papers were single-authored:
http://arxiv.org/abs/hep-th/9510017
http://arxiv.org/abs/hep-th/9711200
http://arxiv.org/abs/quant-ph/9707021

But could you really argue that those were more important than say:
http://garfield.library.upenn.edu/classics1993/A1993LJ13600001.pdf
http://arxiv.org/abs/hep-th/9601029
or the BCS paper?

 

[moontiger]

I don't know about experiment, but yes you can be a solo theoretical physicist and I know a handful of people who are, though usually they have done non-solo training to the postdoc level first.

But it is an uphill battle. The main problem is that without affiliation to a university or other institution you do not have access to most physics journals, computational software, etc.--a lot of expensive resources that are necessary to do theoretical research these days and that people at research institutions take for granted. You also don't have a circle of people around you to discuss with--or cheering you on--so it's hard to stay motivated. Plus unless you are in academia, there is no real reward for publishing...it's probably not going to advance whatever job you are doing 40 hours a week to support yourself. And at some point in life most people acquire spouses and families, and at that point would rather invest their free time there.

I think that's why solo physics doesn't work out for most people.

 

[jack476]

I see. So if someone doesn't like working with others they should not be a physicist?

Working with others is just going to be an unavoidable part of any professional career. There are very few jobs, both in and outside of STEM, that come with the expectation that they can or should be done alone.

Also, why? What has changed about physics that has made it so that someone can't do anything by themselves?

Cost and complexity. Research is an expensive endeavor and it requires the backing of an organization. You also have to work with experts in other fields, for example, modern theoretical physics often makes use of computer simulations which can require a lot of power, so you have to have programmers who know how to maintain and operate, and in many cases build from scratch, a supercomputer. You may require specialized equipment, the design of which will require the input of an engineer. One person can't reasonably be expected to handle all of that alone.

Staying competitive and relevant also means communicating with others in your own field, networking, going to conferences, stuff like that.

Finally, and perhaps most importantly, there is the expectation that you have someone who can check your work. Without this, you risk corruption and turning into a crank.

 

[atyy]

Here is a way to get the best of both worlds - unwittingly collaborate, then write solo papers

http://www.math.columbia.edu/~goldfeld/ErdosSelbergDispute.pdf

 

[jesse73]

Mathematics is different and is funded with a different culture. People in physics in the US are only going to shift more towards working in bigger groups because the funding decision makers at NSF want to fund larger groups which are ideally interdisciplinary. You can do whatever you want in this world if you pay for it yourself.

 

[Ben Niehoff]

Do you know Juan? It's hard to say that he works by himself. He has 14 papers with 500+ citations. Eleven of them have co-authors.

I know Juan, a little bit anyway. He absolutely does not work in a vacuum. No one does.

A single person working alone cannot possibly keep up with what's happening in the field. You at least need to be at an institution where you can talk to people. Aside from that, collaborators bring useful skill sets and knowledge to the table that you may not possesses yourself. And for long, intricate calculations, it's useful to have a second pair of eyes to make sure there are no mistakes.

If Einstein had collaborated with, say, Hilbert, then we would have celebrated the centennial of general relativity ten years ago.

 

[PeterDonis]

If Einstein had collaborated with, say, Hilbert, then we would have celebrated the centennial of general relativity ten years ago.

I think that's a bit extreme. First of all, Einstein did collaborate with Hilbert in 1915; so at best you can say that if he had done so sooner, he might have reached the final form of the field equation sooner. (IIRC Einstein visited Hilbert in Gottingen around June 1915, and Hilbert completed his derivation of the field equation using the principle of least action in November 1915, something like 5 days before Einstein reached it independently.)

But how much sooner? I don't think it could have been before 1908 at the earliest, because the two key physical insights were in 1907 and 1908--Einstein's realization that a body falling freely will not feel its own weight, and Minkowski's introduction of spacetime. And even then I'm not sure the field equation could have been reached right away.

Also, Einstein was by no means the only one working on a relativistic theory of gravity in the period 1908-1915, and he was not out of touch with others who were doing so (Nordstrom, for example).

 

[Ben Niehoff]

I think that's a bit extreme. First of all, Einstein did collaborate with Hilbert in 1915; so at best you can say that if he had done so sooner, he might have reached the final form of the field equation sooner. (IIRC Einstein visited Hilbert in Gottingen around June 1915, and Hilbert completed his derivation of the field equation using the principle of least action in November 1915, something like 5 days before Einstein reached it independently.)

But how much sooner? I don't think it could have been before 1908 at the earliest, because the two key physical insights were in 1907 and 1908--Einstein's realization that a body falling freely will not feel its own weight, and Minkowski's introduction of spacetime. And even then I'm not sure the field equation could have been reached right away.

Also, Einstein was by no means the only one working on a relativistic theory of gravity in the period 1908-1915, and he was not out of touch with others who were doing so (Nordstrom, for example).

My point is that Einstein spent probably 10 years stumbling around in the dark, struggling with things that were fairly obvious to Hilbert. Off and on throughout that time, there are even talks Einstein gave where he insisted that general covariance was not important, and that probably no generally-covariant field equations could be written down that were consistent with nature.

 

[Redbelly98]

Just to clarify, when I say lone physicist I mean someone working on the actual theory by themselves, ie not in a group research project.

One thing that is not clear to me: do you have in mind:

• Somebody with a position at a research or university institution
  
• Somebody with no such position, but has a degree in physics
• Somebody without a degree or formal training in physics

What you are asking gets progressively harder as you go down that list, and many would argue becomes impossible (at least for practical purposes) by the 2nd or 3rd bullet point.

 

[Vanadium 50]

I am surprised that this has gone on as long as it has. People are having to look at centrury-old cases to see if they might be an example. That says something, no?

 

[Evo]

This is long done.