Among users of Physics SE there is a group of people who wants to have their 'Physicsoverflow', that is, a group oriented for research-level questions.

It is almost impossible to turn this SE into research level, for three main reasons:

  • The Physics SE was founded with purpose for general audience (see definition phase of Physics SE).
  • Simple (but not stupid) questions in physics are important (I never tried to discourage asking or answering them).
  • Now there is only a tiny fraction of research-level questions.

There is proposal of Theoretical Physics SE with ambition to become 'physical Mathoverflow'. I like the idea very much, but with one (albeit strong) reservation: should it be restricted only to theoretical physics? I would much more like to participate in Physics Research SE, because:

  1. Most conferences I attended (in wave and quantum optics) had both theoretical and experimental lectures&posters and it was stimulating.
  2. When asking research-level questions both theoretical and experimental answer would be appreciated (at least, for me).
  3. Good experimentalist need to know theory. Physics is immanently connected to experiment.
  4. Physics - pure maths - experiment = applied maths without an application.
  5. I am not afraid of abundance of experimentalists as it is somehow easier to ask theoretical question ("what is wrong with my assumptions" vs "what is wrong with my experimental setup").
  6. Anyway, most of the users will be interested only in some questions. Even a factor of two won't break anything (and you can filter experimental questions if you really want).
  7. We need people to join - I doubt if there are by now too many physicist for the one research-level SE group.

However, I do not know what percent of potential users of Theoretical Physics SE share similar attitude. Also, I think it is not the best to propose another SE without a proper consultation.

Relevant discussions:

The main question are:

  • Who of you are interested in having general Physics Research SE rather than Theoretical Physics SE?
  • Are on Physics SE any experimentalists wanting to ask&answer research-level questions?
  • In your opinion, should I propose Physics Research SE?
This seems like the kind of thing that could be discussed on the new site proposal page on Area 51. – David Z Nov 15 '10 at 22:22

10 Answers 10

I don't think another site is a bad idea, but I still don't understand why you guys do not ask what you consider "appropriate level" on this site.

All this thing is based on the assumption that things cannot be changed.

That being said, the proposal is again based on another assumption: theory + experiment is too broad.

So what will be the next attempt to divide the community even more ? Apparently PSE spent a very long time before the beta, maybe we should try to understand why before proposing one or two more.

Nothing personal, nothing again your proposals, just my opinion. – Cedric H. Nov 15 '10 at 18:26
My point was to divide graduate-and-above from popular&undergraduate (and make no further divisions). It's not only the level here (actually, it is becoming better, with increased amount of good undergrad questions). The thing is to get the critical mass of researchers you need also to have high density (see:… - based on experience of Mathoverflow and CSTheory). Otherwise (i.e. without dividing) - Physics SE will have some fraction of research questions, but it never be even remotely close to MO or CST. – Piotr Migdal Nov 16 '10 at 10:58
+1. Too much division only serves to weaken a community. I think Physics.SE needs to grow far more before considering any need to split it. Also recently it has finally begun to attract high quality people in large numbers and a correspondingly large number of high-quality of questions and answers so I'm very optimistic about its future. I'm not even sure why the separate theoreticalphysics.SE is needed. That is what tags are for. – Deepak Vaid Jan 20 '11 at 12:46

I don't think it is a good idea; SE site needs mass (both in users and questions/answers) to be self-sustained; without it, it would become a chat place for a few people. Also many sites on a similar topic makes a migration hell and some inconvenience in using few sites at a time.
On the other hand, I don't see a reason why mixing trivial and complex questions may be harmful; SO is doing very well with both questions about merging strings in JS and optimizing Erlang code. Not to mention that sometimes it is harder to explain something plain than to cite something complex.
Broad scope is also not a problem; first of all one can always use tags to filter questions, second mixing fields tend to a creation of novel views on a problem. CV evolved from Q&A for statistics to Q&A for statistics, data mining, machine learning, visulisation and data-driven computing and nothing bad happened, even more, some communities stuck in Area started avid activity there.


I don't really see any reason to split the site. StackOverflow has somewhere around milion questions at all levels from basic to very advanced and only recently it split.

This site has not even left beta and already it feels like splitting. Why that? Are you annoyed by stupid questions? Skip them! Are the quetions to simple and not intriguing enough? Spliting will not bring more or more interesting ones.

Have a theoretical-physics and research-level tags and filter the other questions. Otherwise it might end with one site with 20 bored PhD physicists with no questions and 200 students and enthusiasts with lots of questions but no answers.

Recently I got involved in leaking wagon discussion and even the top rated users still do not have a clear answer everyone would agree on (at least I cannot find it there)... And it would not turn better with less users. You don't raise the level of the site by deleting easy questions, bud by answering them correctly. The site needs some basement to grow on.

You make very fine points, +1! Still, I think dichotomy like Mathoverflow/math.SE can work (and does work) very well. And I don't know about SO... is an actual research being done there? If so, why was cstheory.SE created? – Marek Dec 13 '10 at 23:32
I am not sure why cstheory has been created, but probably because SO became too big, there are more programming related sites (SO, SF, cstheory, programmers + a few smaller). I do not say it should never split, but it should split because its too big, not too small. – Lukas Dec 13 '10 at 23:47
well yeah. But I don't think anyone is splitting here. There is some overlap between physics.SE and TP.SE (e.g. me) but I won't leave this site just because a new site is created. I'll ask my research level questions there and help to solve lower level questions here. I honestly don't see any problem with this. – Marek Dec 14 '10 at 8:51

I agree with your questions, but also struggle with the precise name. "Theoretical/Experimental"? Maybe just, "Research Level Physics SE"? The current proposal is too good to waste, however, so maybe just start by proposing a few good experimental questions there.

Yeah, experimental is an important part too. – Noldorin Nov 15 '10 at 15:27
When it comes to name, I am very open to discussion. The only thing is that I don't want to have it too long. – Piotr Migdal Nov 15 '10 at 15:32
I like "Physics Research SE". – nibot Nov 15 '10 at 18:58
Kind of analogous to the usenet group sci.physics.research. – nibot Nov 15 '10 at 18:59
If we get one or two experimental question(s) upvoted as on-topic, on the TP.SE, the name can be changed. I guess it's the ideal way to decide. – Frédéric Grosshans Nov 18 '10 at 17:39
It should be obvious that explicitly experimental question don't belong on a site named "Theoretical Physics", and it is certainly the case that there are "research level" questions to be asked about experiment. So either a name change would be needed or we'd need to get started on a second "research level" physics site. And then some smartarse will ask where phenomenology questions go... – dmckee Dec 4 '10 at 0:40

Since there is a now a discussion site at area51 I posted one there with the following suggestion:

As long as theoretical physics site does not even exist, ask your toughest questions at physics.SE. That way you gain rep to help progressing commitment phase for one thing, and, what is more important, you help supporting physics.SE with high level questions. In the hopefully occurring event that both sites make it out of beta, questions can still be migrated at a later point (and merged if cross-posted to seed TP.SE beta).


I've changed my mind completely on this. I don't think you should split the site. I can navigate quite reasonably with the tags. And right now, the site is WAY TOO SMALL to split.

When the number of questions becomes so large that I'm unable to read the day's new questions in 15 minutes it might be time to consider splitting, or some way of allowing a reader to get only a partial sample of the questions. For example, if all questions were required to carry one or more of three tags: "Experimental Research", "Theoretical Research", or "Not Research", or something like that, then those only interested in a single factor could read that part.

Previously I had written the following which I will leave here for continuity:

You absolutely should split the site. The only reason I'm hanging around is because I've got time on my hands. That will end in the next few weeks. If there were a research level place I'd be happy to contribute, but to put research level questions here is a waste of time.

The purpose (to a researcher) for having research level questions is to widen the number of people who are aware of a question. The problem with this site is that the level of the discourse is too low.

One thing that might help with a research level site is to require that the posters be actual researchers. A good requirement would be at least one peer reviewed article published in the past 2 years.

I'd +1 you except for your last point. The community will sort this out by voting and closing and if necessary banning, no need to require an artificial definition of "actual researcher" - e.g. what about PhD students that have just started? – Tobias Kienzler Jan 20 '11 at 9:43
I largely agree. However, you don't need a minimum competency requirement. In practice once you get in enough competant people at the start the site becomes self policing (see CSTheory and MathOverflow as examples). – Joe Fitzsimons Jan 20 '11 at 9:56
Dear @Carl your last suggestion sounds odd especially coming from someone who has themselves done an impressive amount of work outside the academic mainstream. – Deepak Vaid Jan 20 '11 at 12:41
Tobias,Joe: I doubt that there will be enough cream to change the character of the milk. I don't know anything about CSTheory or MathOverflow; maybe we're not doomed. Space_cadet: Note that I didn't say "must be in academic mainstream", but instead publishes papers. And this is just from my experience. I'm having a lot of fun posing phase problems and right now I think the system is working well. – Carl Brannen Jan 21 '11 at 4:40
Okay, I admit I was wrong here. I'd forgotten I'd even written this. – Carl Brannen Jan 30 '11 at 0:47

One thing you should always keep in mind: Jeff et al. might at one point decide that one physics.SE is enough, so you shouldn't assume PR.SE will definitely be there one day. But if there will be one, there will almost certainly be a question migration path to here and vice versa. So why not ask research level questions here and if PR.SE starts have them migrated or ask again?

There are much more people interested in popular-level than in reseach-level physics. I don't say I want to leave this site - just I know that without research-only group it is not possible to get critical mass of researchers (not only - popularizers). See:… While some of the questions may lay in the intersection (i.e. (advanced undergraduate / graduate) questions that happens to be answered in a textbook), I believe this intersection isn't that large. – Piotr Migdal Nov 16 '10 at 10:36

I suggested the focus of the proposed site be on theory simply because when you combine both theory and experiment I suspect you get a field so broad that we would struggle to get consistently good answers. I would suggest that if you want to have an experimental physics site, that someone propose it, and at a later stage it is possible that the two sites would decide to merge (a process which has been done before on SE). However, for the moment I would suggest maintaining the restriction to questions with a theoretical bent.

There is huge intersection of theory and experiment. Even if I am interested in theoretical part, IMHO the intersection is crucial. Words "Don't even try to calculate it, because in real materials the assumptions (a) and (c) are never fulfilled at the same time" may save months of futile work. If we are not sure if there is a critical mass of potential users, IMO much better is to create one research-level group and then divide (in case of 2 non-intersecting themes) than to create two and risk failing both + never learn if they can interact. – Piotr Migdal Nov 15 '10 at 16:10
@Piotr: I think you are perhaps not giving the communities enough credit. Experimentalists and theorists interact all the time. I myself have been an author on a number of largely experimental papers. I simply think that making the scope to broad is a bad idea, as is making it to narrow. – Joe Fitzsimons Nov 15 '10 at 16:44
I posted a question example "Does Brazil nut effect depend on coefficients of friction and resitution? (theoretical, numerical and experimental answers appreciated)" (well, the bracket is there only to make the question explicit). If there where distinct 'Theoretical Physics SE', 'Experimental Physics SE' and 'Numerical Physics SE' then I had to send it three times as (perhaps) a huge share of other questions. – Piotr Migdal Nov 15 '10 at 17:11

I saw this thread and (existence of "theoretical" in Area 51) only some days ago. I regret that very much, because this thread explains a lot of things I did not understand. I am not qualified to write in there, but I will read and hopefully learn something. This was my original intention here too, and I was very surprised that I could write here answers or correct even moderators in some cases. I learned too, but that was rare.

Now with the thread above I understand more of the original intentions of this forum and the reality which took place. (And why I was silly to fight that). The "poll" above, including that of the two or three main advocates of mediocrity for allowing low leves questions, made everything clear. Good luck.

I hope "Good luck" doesn't mean goodbye – Tobias Kienzler Jun 21 '11 at 8:13
@Tobias, no :=) This was meant as good luck to the forum and its "matadors". – Georg Jun 21 '11 at 9:18

It seems important not to be afraid of this good development. Math.Stackexchange continues to have many interesting questions and answers. Two sites are pretty clearly needed, so that there is someplace to refer the general questions. It's just that in this case the general site was founded first.

Your post seems vague to me. By two sites you mean (Physics SE + TP/PR SE) or (Physics Research SE + Theoretical Physics SE) or something different? – Piotr Migdal Nov 15 '10 at 16:15
Phys SE +TP/PR SE – sigoldberg1 Nov 15 '10 at 19:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .