MathOverflow policymaking and transparency

Question

There goeth rumours abroad that some off-site groups are seeking a private forum together with the moderators here in order to make policy for MO.

I hope that the moderators will commit to transparency with regard to policymaking and host any such discussion here on meta or in a place where it can be monitored and commented upon by the community. I have no reason to doubt that the moderators would indeed demand that such a discussion take place where the community can participate, but it would set my mind at ease if the moderators would agree to this explicitly.

Answer 1

The rumor "some off-site groups are seeking a private forum together with the moderators here in order to make policy for MO" is incorrect, in fact way off the mark. There aren't off-site groups knocking on the moderators' door, much less seeking to help set policy for MO. That sounds like the bubblings of some conspiracy theory.

I will attempt to clarify what is really going on, before more such "telephone" rumors emerge.

There is widespread perception within the mathematics community at large that MO is important but can be a highly unwelcoming place. You may disagree with that, but the perception is undeniably there. And of course the members of the MathOverflow board are highly aware of that perception, and that it may be a serious problem for us in the long run. What should or can we do about it, if anything? We really don't know. Should we simply close our eyes to it, pretend there's nothing there? Doesn't seem like a good idea.

After some internal discussion within the board (which consists of the current moderators, Anton Geraschenko, and Ravi Vakil who has been part of the MO venture since the very beginning), we've put together a list of people who we thought might be able to help clarify the nature of the perceived problem. The list comprises both MO community members of long standing and respect (both here and IRL), as well as others outside who are prominent in the mathematical community, and who see MO from the outside. It's we, the board, who are seeking them out -- not the other way around. Only a few so far have gotten back to us. And they are not 'groups'. They are disparate individuals that some of us know, who have their own ideas and opinions, and who will probably disagree with each other in various respects. They didn't even know this was coming. But they are on the list because at least some of us know them, and trust to be people who can help drive a fruitful conversation.

And for now we are simply proposing to have a discussion (none has occurred yet), and yes, if we reach some shared idea about what the problem is or how extensive it is, we would also encourage suggestions for what (if anything) might reasonably be done about it. This is not even remotely close to the point of "setting policy"! For example, there would be no 'vote' taken within that group. We would not propose or presume any policy-making without an open discussion, here, before the community.

The list of people we compiled is not a matter for the MO Community to vote on, nor is it even a matter up for discussion. It is not after all a deliberative body that will vote to make decisions for the MO community and for which minutes will be taken and made public. Additionally, I hope it makes sense that sometimes, in order to have a frank and productive conversation where people can freely speak their minds, it's not always a great idea to have everything be public. In the present case, we feel discussion will flow more freely if the participants know that their speech will not be dissected in a public arena so that every word must be carefully weighed and measured.

Answer 2

I thought I should respond too, to follow up to @Todd Trimble's reply, to try to be as clear as I think reasonable.

This is not a board initiative, or an initiative of the moderators, although it arose out of many discussions of many different people over many years, including board discussions. But the board is aware of the initiative, and I believe the members of the board are individually supportive of it (although there is no vote). I understand that there is not universal agreement that MO is perceived as unwelcoming, especially to certain groups of people, and there is also not universal agreement that for the good of mathematics, MO should be more welcoming, and if so, what things can be influenced. I wanted to ask some people to have an informal discussion about it in one place (rather than serial small discussions over many years), and I picked them quickly and without much thought (because I wanted to get it done, and I have limited bandwidth, to be honest). I picked people who I think will be able to have a productive conversation together. I'm intending to have this discussion over the course of a month. This entire "group" is not official, and there is no way for it to set policy. But I want people to speak frankly, and not be reluctant to speak because they feel they have to choose their words carefully. So I've asked people to keep the conversation itself privileged, and I also want them not to share the names of the other participants (although they can of course say they were part of the discussion, and talk about it in general terms). For the discussion to work, there needs to be a high level of mutual trust. I knew in advance (and the board did too) that this could be taken as some secret cabal of highly-connected insiders. All I can I say is that this isn't the intent, and so if it is a secret cabal, I hope it is taken as my own action, asking people I know if we can chat about what I see as an important challenge for MO. And so the full blame should fall on me for asking people, and not on anyone else (including any of the moderators, as they did not officially sanction this in any way). For what it is worth, I invited all the MO moderators, and roughly half (I haven't checked precise numbers) are taking part. (Interesting that news of my effort leaked out before I'd even invited everyone! I will actually be a bit cautious with the rest of the invitations, because I don't want to hurt feelings.)

If there are things any of you want to pass on to me about this, please feel free to email me, and I will intend to pass everything on to the group (unless you tell me otherwise). If there are things any of you want to say publicly about this, then please do so (here on meta). If there are discussions you wish to have in other venues, of course by all means do, and I encourage you to do so. (Please be aware for both inflow and outflow reasons, I am very slow to reply to emails. I wish it weren't so. But this doesn't mean they don't get read.)

Answer 3

I personally trust the judgement of the moderators, and would not want to prevent them from having some preliminary stage-setting discussions with other parties in private if they felt that that was appropriate. Obviously any actual change in policy will need to be discussed openly here.

Anecdotally, it seems that there is a substantial class of potentially valuable contributors who feel that the basic framework of MathOverflow is unwelcoming. Whether or not you agree with that, it seems suboptimal to force them to use the same framework for even the earliest stages of a discussion of their concerns.

[UPDATE] Let me amplify my thinking a little. I am guessing that the mooted discussions mostly arise from concerns about gender balance. It is undeniable that among users of MathOverflow, there are only a tiny minority who are women using their real names. There may be a few more pseudonymous women, but nonetheless it is clear that the proportion of women on MathOverflow is far below the proportion of women in academic mathematics. I find this disappointing; there are certainly very many women whose expertise and insight would be extremely valuable if they chose to participate. I have not seen an explanation of this phenomenon that seems compatible with easily obtainable data.

Of course, we should consider explanations like "women do not regard MO as a productive use of their time, and they are correct in this judgement". While this should not be rejected out of hand, it would be strange to have a huge gender imbalance in judgements about the value of MO activity, and this would itself seem to need explanation. Often I see people suggesting that women are unwilling to participate in some kind of forum because they suspect that they will be harassed, or their qualifications belittled and their opinions ignored. I have tried to search for instances of this sort of thing on MO, and have found very few. In cases where women do participate using their real names, it seems to me that interactions are very normal, with other users being respectful and appreciative to experts, and helpful to novices. Certainly I see MO as being much more friendly to women than Twitter, despite the fact that women are far more common there.

This situation seems quite mysterious to me. I think that it is valuable to seek insight and suggestions from as many places as possible, even if the community ultimately decides not to accept such suggestions. So if there is some group who may have useful ideas who are reaching out to the moderators, then I think that the moderators should be as receptive as possible without committing the community to any course of action. Because I think that the situation is mysterious, I think we should avoid making assumptions about what kind of insights and suggestions we might receive.

Answer 4

A reasonable process for new "policies" in my view is:

1. Moderators and the MO board (or, for that matter, anyone else) discuss and make a proposal.
2. This proposal is discussed on Meta, so that reasonable changes can be made, and users can express their disagreement.
3. The (amended) proposal is implemented and made effective, if the board sees it fit.

If Harry is suggesting "let's not skip #2", I agree 100%. If Harry is suggesting "let's skip #1" (as it seems now), I do not agree. It is unreasonable to forbid the MO board to have private brainstorming sessions. Should we also ban them from discussing MO in the coffee room?

Let's not forget that the MO board has ultimate control on this platform, in the end, while we users have the freedom to leave and fork it if we don't like the changes.

Answer 5

I have absolutely no knowledge or involvement with any matter involving MathOverflow governance, including the initiative described in this question. I'm commenting based on my experience with policymaking more generally, both as a creator and subject of policy.

My opinion, in short, is: transparency at the early stages of the policymaking process is in nobody's best interests. The basic reason is that the process works best when many ideas are considered but the bad ones are killed off as quickly and secretly as possible.

Suppose it is discovered, for example, that a number of MO users who post about differential geometry a lot are toxic and driving away users that MO wants to attract. (This example is purely fictional.) Suppose somebody proposes to deal with this problem by banning the differential geometry tag from MO, almost certainly a bad idea.

Scenario 1: public debate. Yes, there would be good arguments about how it's silly to ban an entire field of mathematics due to a few bad actors. But there would probably also be bad arguments made by people who don't like differential geometry, and the toxic users themselves would probably make a scene. A big mess would ensue, and the whole episode would risk driving people away from the site.

Scenario 2: private debate. Even if the proposal had one or two passionate advocates, the argument that banning a major field of mathematical research will help attract more MO users is untenable and would be killed quietly, assuming a suitably informed and diverse group of participants are invited to the discussion.

Obviously scenario 2 is better. Out of the box thinking is good, but some ideas should be quietly put back in the box.

Of course, the extreme lack of transparency also carries risks - I wouldn't want MO to unveil a bunch of controversial new policies without providing users with a chance to ask questions and propose changes. Though the MO board has every right to take such an action if deemed appropriate. If the site had significant financial incentives to aggressively drive user growth or appease a corporate investor then I might be concerned, but so far as I am aware everyone is properly incentivized to create a healthy community for mathematicians.