But since I never ask anything it is not a personal concern for me, and being pseudonymous and all I might not be well placed, regardless. It was just surprising for me, so thanks again for the elaboration. Certainly one can have quite differing and equally valid points of view on this matter.

The original author still owns the copyright and they are free to do whatever they wish with the original work. You own your work always and forever under copyright law. We have not nor can not take that away from you.

Take a look at the below from the Math meta site and let me know if you have any further questions.

How can I make my contributions to Public Domain under my specified license in Account Deletion?

- Stack Exchange Team

(However, I don't think that post on meta.math.stackexchange.com has any legal weight as a clarification of the SE TOS.)

I thought (and still think, but I'm not a lawyer) that if you license content to MO, there's nothing stopping you from re-licensing it elsewhere under different terms. In particular, I don't think content authors are subject to the constraints of the attribution page.

I certainly think this is how it should work, and the CC BY-SA license in no way prohibits other licenses in addition, but it would be best to clarify or revise the terms of service for MO 2.0. The current SE terms of service say explicitly that the specific attribution requirements apply to using your own content and not just other people's. It's not clear to me exactly what this means (one interpretation would be that by using the site you are agreeing to the terms of service and thereby restricting your ability to re-license your own contributions). I assume this doesn't really matter to the SE team, in which case changing the terms of service should be fine, but it's a little worrisome that it apparently mattered enough for someone to insert the "whether such Subscriber Content was created by You or others" clause. It certainly looks like the terms of service are trying to forbid re-using your own contributions without crediting the stackexchange site, and the only legal question is whether they have left loopholes.

(MO as coauthor) [...] I don't understand why the legal entity MO should have any opinion about this or take any action with regard to it.

The issue, I think, is that permission to use the MathOverflow name is now (legally) necessary and therefore somebody may ask for it. Do we need to prepare for this? The ethics are a non-issue but the laws can be tricky. My conclusions were discussed above; anyone is welcome to do more research about this. There are tons of little things to worry about here, especially in connection with 501c3 requirements which put some restrictions on what MathOverflow can do.

Common sense seems to work well for now. I don't think we need to do anything until somebody asks. We are protected to a limited extent from people using the MathOverflow name in illegitimate ways. (For example, I'm pretty sure we won't get in trouble with the IRS if someone else makes a political campaign contribution under the name MathOverflow.) As far as I can tell, there is no need for additional protection of the MathOverflow name (and I don't know why there would be interest in doing that). We also should avoid giving blanket authorizations to use the MathOverflow name or make vague promises in that direction unless we understand the ramifications. (That's a "rule" for board members, not for community members.)

If you want to use the MathOverflow name for something, just ask! We'll check if that puts the company in legal trouble and grant permission accordingly.

Also, I have nothing against, it could be even very good, some compilation of MO content, I only mentioned the combination of it pretending to be "official" and at the same time being really bad as a problem.

(MO as coauthor) Please don't make MO (or "Matthew O. Verflow") a coauthor on anything. It makes about as much sense as making Hotmail or your university common room into a coauthor. MO doesn't do math; it just connects people so that they can do math together more effectively. Then again, if you want to be silly like that, I won't stand in your way even if I find it annoying. I don't understand why the legal entity MO should have any opinion about this or take any action with regard to it.

(MO copyright) Again, I don't think the legal entity MO should do anything. When MO started, I basically copied whatever Stack Overflow had as their attribution agreement. The intention was to prevent stupid stuff like mirror sites which just copy content. I thought (and still think, but I'm not a lawyer) that if you license content to MO, there's nothing stopping you from re-licensing it elsewhere under different terms. In particular, I don't think content authors are subject to the constraints of the attribution page. Anyway, even if you copy a thread verbatim (which you wouldn't do in a paper anyway ... you'd just express the ideas, so almost all such copyright discussion is moot), it would still probably be fair use. (IamnotalawyerIamnotalawyerIamnotalawyer)

(MO trademark) I don't think the legal entity MO should do anything, except perhaps to prevent itself from being bullied by trolls (which I think is what François was looking into). But I don't think we should expend any more effort than is reasonable. I find it very difficult to imagine a situation where we (the MO community) lose the ability to use the domain mathoverflow.net, even if some other entity decides to call itself "MathOverflow" and threaten us that we'd better not use the same name. Maybe they even manage to legally prevent us from using the MO logo. In that situation, it seems pretty painless to just keep doing the same thing at the same url with a different official name and logo. Who cares? If we're really worried that some bozo is going to start making "official" MO stuff that we don't like, we'd be much better off making a big joke of it by encouraging everybody to make their own "official MO" junk, with extra points going to whoever makes the stupidest, most knock-offish thing that makes fun of the bozo.

To recap, I think the MO legal entity should be a little-used tool of the mathematical community. It should only act when directly motivated by the goal of helping people do math together, and when the MO community (or somebody else) cannot just as easily accomplish the same goal without it.

If you want to share credit with anonymous users who don't want to be co-authors or can't be contacted, you can use the standard method of adding a line in the acknowledgements, e.g., "MathOverflow user 11961 ``Donald Duck'' contributed substantially to this project, but could not be contacted/declined co-authorship."

At this point, MathOverflow is incorporated but it's just a bare bones company. (By the way, MathOverflow is the full name, no Inc., LLC, or whatnot.) The idea is to get something set up quickly so we can sign a contract with StackExchange and migrate to the network. Our lawyers are currently working on drafting bylaws and basic structure and filing a 501c3 nonprofit application with the IRS. That will take a long while and our immediate concern is just to avoid screwing up. We want to get enough structure that MO is protected and remains at the service of the community.

I don't think much of what we set up now will be permanent, it's intended as a transitional structure so we have the authority to sign contracts and secure funding for MO. None of us are interested in running the place that way. From the inside, it's a very strange experience, we're always a bit on edge when dealing with all this since we only know what things really mean until much later in the game. For example, we don't really know when and why MO's physical address will become relevant but we can't wait until then to pick one. But there is a good side to all this: we're learning a lot about the various components necessary to make MO work and we will be able to make much better decisions about the future of MO once we get through this.

Things will work out in the end and our priority is to make sure that MO the community does not suffer through the transition.

Rather it seems to me since, as far as I understand, he specifically is at the moment busy with some administrative/legal work in order to ensure that MO can continue well these types of questions are at the moment on his mind and thus when the suggestion came up "use MathOverlow" it occured to him that there is or at least might be a general question to be dealt with who can use the name MathOverflow for what purpose (as the situation everybody, and I mean strictly everybody, can use it for strictly everything could create problems).

I will say when I first read asking board of directors for permission and so on I was momentarily suprised, too, as it sounded strange, but then this board is nothing 'mysterious' but as mentioned recently in the migration thread it is the moderators plus Ravi Vakil. So, in the end it is anyway what seems common sense-ish to me, if you want to do something non-standard relative to MO you will bring up the question here or with the moderators directly, and then something will be figured out.

And, I have full confidence the moderators have no interest at all to impede reasonable activities by users also involving the name MathOverflow or will start creating problems for people when attributions guidelines are only followed in spirit but not by the letter (due to loversight or lazyness).

But, there could also be most unreasonable activities by non-users or also certain users (I mean there are also some users on the crankish side around here). Say, I would not like to see, and it could be harmful, a book or website pretending to be some "official MO Best Of" while being some badly produce compilation made by some unqualified person.

Or, there could be still other things, say somebody thinks it would be a great business idea to spam MO-users and/or other mathematicians with promotion for whatever math-related stuff under the name MathOverflow.

Then it could be good to have some grounds for saying this should stop. So, it seems reasonable and even desirable to me to think about such things.

To turn this around if person Q asks a question and person A answers then also person Q could consider using the answer of person A in a publication (with attribution). Yet, if then both should happen without this being checked in advanced this could cause problems. Indeed, I consider this second scenario as least as natural. If I give an answer (without saying otherwise) the person I gave the answer to can of course use it for whatever they want to use it; that was the point of answering it after all.

Thus, in my opinion, as a general principle (there can be unusual circumstances) a person answering a question and wanting to publish the answer should at least get the consent/opinion of OP of the question.

Of course, common sense would suggest that one say cannot create another math Q&A site taking all the existing content of MO as a start, also call it MathOverflow, and host it on mathoverflow.[sometopleveldomainthatisavailable] but then perhaps somebody might consider it. What would you suggest doing in this event? Just ignore it? It could cause some confusion after all.

At last, what exactly are you afraid of?

That was the question, fedja. A particular case was brought up by Gerhard where someone used MathOverflow as an author of a paper. I investigated these issues and I got most of the answers I wanted. Thanks for your help.

First, there is a big difference between a corporate name (like the one we have: MathOverflow) and a trademark (which we don't have). The latter is an actual property which reserves exclusive use of the name/logo for various purposes. It doesn't look like we really need a trademark since we have earned the right to keep using the name for the purpose of this site. However, there is a lot of "advice" out there claiming that these rights are not sufficient protection. We do have a clause in our contract preventing SE to use MathOverflow for another similar site and that's probably the only thing we really need to worry about.

By incorporating in Delaware, we ensured that MathOverflow is a unique identifier in that state, some other jurisdictions would not allow the reuse of that name by others but Delaware is the only sure thing. The name MathOverflow is still somewhat protected against misrepresentation in most jurisdictions, nationally and internationally. In other words, it is illegal to use the name MathOverflow, in reference to this corporation, without permission. There are tons of loopholes such as using MathOverflow in reference to something else (e.g. a corporation with the same name in another jurisdiction) but it is still unlawful to do that with the intent of deceit. Bottom line: you can't use MathOverflow as an author of a paper without permission unless you're referring to something other than this site.

This mention of the by-laws made me think of you as Dennis, and the autonomous collective. (Note that Dennis talks about classes, much like set theorists!)

For the sake of clarity, what I had planned to bring up to the board is whether there should be a policy about the use of the name MathOverflow, which includes but is not limited to this particular question of authorship. I'm checking what legal issues there may be and whether there is a basis for protecting the use of the name MathOverflow. We might as well take care of this now while we have our lawyers on a retainer.

I agree with Henry and Scott that the wording of the TOS is not as clear as it should be, we should tell some SE folks. The reading that requires the that You must follow the CC BY-SA attribution rules even in situations in which those rules would not otherwise apply appears to remove some of the author's copyrights. I'm pretty sure that's not what SE had in mind. The key to understanding the TOS may be the precise definition of "You" and "Subscriber Content" but that is definitely not what I would call clear.

My reading is that the TOS say that even the contributor must follow the attribution rules when posting their content elsewhere. This isn't required by CC BY-SA itself.

In particular, I wasn't too thrilled to see François' suggestion that he would bring the question of authorship by "MathOverflow" up at a board meeting. What's wrong with hashing it out here?

I think we should aim to preserve as much as possible the current governing model. (Or perhaps the governing model that I've had in mind, whether or not it really exists!) That is, to the extent possible decisions should be made through argument (and hopefully eventual consensus) between everyone concerned, here on meta (or its future replacements), with the role of the moderators ideally restricted as much as possible to emergency operations (suspend this person, write to that person privately when a private intervention seems best) and to "infrastructure" work (e.g. setting up LaTeX, citations, arxiv trackbacks, and the migration) which needs continuous attention.

To the extent that for legal purposes decisions are required of the board, I hope that as often as possible these decisions are merely rubber-stampings of decisions reached amongst the wider MO community. That said, I'm pretty content with the moderators (and hence the board) making their own decisions when consensus here on meta is not achievable.

On a purely selfish note (as someone on the new board of directors), I hope that our board meetings (note that none have taken place yet) are as swift and painless as possible, because everything that possibly could be taken care of here on meta already has been!

Incidentally, for me the weirdest and most confusing part of the terms of service is the following sentence:

In the event that You post or otherwise use Subscriber Content outside of the Network or Services, whether such Subscriber Content was created by You or others, You agree that You will follow the attribution rules of the Creative Commons Attribution Share Alike license as follows:

I don't understand what it means to say you "will follow the attribution rules ... as follows"? Does it mean when the attribution rules apply, this is how you should follow them? I.e., in this case it's just explaining what you have to do to "attribute the work in the manner specified by the author or licensor" as required by the CC BY-SA license. Or does it mean you must follow the CC BY-SA attribution rules even in situations in which those rules would not otherwise apply? As the copyright holder, you can normally do whatever you want with your own work, including using or releasing it under other licenses, but this interpretation would forbid that. I can't come up with a compelling argument for why either interpretation is untenable, but of course I'm not a lawyer.

In any case, regardless of how the decision to do so is reached, in order to use MathOverflow as an author several things need to happen, including (1) getting permission from the board of directors and either (2a) getting all contributors to relicense their work to MathOverflow in a manner that doesn't require precise attribution or (2b) annotating the paper with precise contributions to fulfill the requirements of CC BY-SA and the SE/MO TOS. Whether or not the board should ever authorize this is the interesting part for me. Fortunately, I don't think we need to seriously entertain this until someone goes through the licensing/attribution hurdles.

I can see MathOverflow being an author for certain sorts of compilations. For example, if someone published a book that collected the most popular MO questions and aswers, then MathOverflow would be an amusing and reasonable author to list for the book as a whole. The CC BY-SA license would allow this, provided the individual questions and answers were correctly attributed and the book itself was released under such a license too.

Henry, I don't agree with your parenthetical remark: when you use a service you implicitly agree to the terms of service. (Note, however, that the SE terms of service essentially prevent giving up claim to individual ownership, in a sense.)

Aside from the issue of anonymous participants, is there a reason why listing MathOverflow as an author would make more sense than listing M. F. Oberwolfach as an author? Polymath-style community efforts seem to me to be rare.

1. Others feeling they weren't given enough credit?
2. A referee or reader thinking that you 'cheated' somehow because the question arose at MO?
3. ...

It seems that 1. is easy enough to deal with: offer everyone involved coauthorship! With careful acknowledgements and context, I doubt 2. is really a problem. There are certainly many instances by now of MO questions resulting in papers --- have a look through the "success stories" thread for some examples.