27
$\begingroup$

In view of recent events, we've been reconsidering having some policy against anonymous users. We've always encouraged users to use real names though we have been tolerant of pseudonymous and anonymous users. As an internal moderation policy, we've always had less tolerance toward anonymous users but we never set any official rules or guidelines. That said, it is becoming increasingly clear that our lack of policy on anonymous users is becoming problematic and encourages uncivil or disruptive behavior in our community.

What anonymity policy should MO have? Should we ban any users that use anonymizing tools to hide their identity? Should we require using real names? Should we maintain the status quo? Are there other things MO could do to encourage civil and responsible behavior for all users?

As announced previously, the moderation team has decided to require certain users to use real names. This is meant as a means for protecting the community as well as people who may be victims of identity hacking. We aim to be fully accountable for this process and provide the community with any information or data that neither compromises the identity of users nor infringes on our privacy policy.

[Edit: We have stopped requiring certain users to use real names.]

$\endgroup$
30
  • 8
    $\begingroup$ Did you ever consider that the real-names policy is part of the problem and not of the solution? To me this seems quite clear. $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 1:33
  • 9
    $\begingroup$ @quid: Could you expand your comment? How might the real-names policy be part of the problem? $\endgroup$
    – Joseph O'Rourke
    Commented Nov 17, 2015 at 1:43
  • 22
    $\begingroup$ It does not seem possible to me to answer the question of what should be our anonymity policy without understanding the nature of the "attacks" and other recent problems, which (for good reason) cannot be discussed. So I am content to trust the moderators to set a policy without much input from the community, which input will necessarily be largely uninformed. $\endgroup$
    – Joseph O'Rourke
    Commented Nov 17, 2015 at 2:28
  • 4
    $\begingroup$ How do we verify that a user is using his/her real name? (For example, what if the user is a new graduate student and doesn't have a department website or any publications?) (I know that this has been discussed many times before, but perhaps someone could provide links to the discussion.) $\endgroup$
    – JRN
    Commented Nov 17, 2015 at 7:22
  • 7
    $\begingroup$ Would simply suspending or removing fresh accounts the moment they start to misbehave be an option to deal with this? Manually confirming the real name of users is a lot of work and the process could be exploited. In cases where problem users try to hide their identity I find it easier to simply act on the base of behaviour instead, then I don't actually have to find out if it is the same user. $\endgroup$
    – user35354
    Commented Nov 17, 2015 at 8:56
  • 6
    $\begingroup$ Isn't similar problematic users present on other sites like MSE and SO? I don't see why MO would need a different way to deal with problematic users. $\endgroup$
    – Kaveh
    Commented Nov 17, 2015 at 21:57
  • 8
    $\begingroup$ @quid If users evade a ban by behaving different enough to avoid being recognized and by not causing the same kind of trouble again they were suspended for, I'd say "mission accomplished". There is not really a way to entirely block a user determined on getting access and a lot of time to waste. Focusing on problematic behaviour protects the site and avoids chasing users that you can't always catch anyway. I like to avoid spending more time catching trolls that it takes them to create new accounts. $\endgroup$
    – user35354
    Commented Nov 17, 2015 at 22:26
  • 4
    $\begingroup$ @Kaveh We've had to deal with anonymizing tools for a while and our defenses there work pretty well. However, such tools are definitely facilitating the problem. I'm sorry I can't go into details right now. Let's avoid playing the 20 questions game. We don't plan on keeping details hidden forever; we just haven't had time to fully assess the situation. I guess I can say that the attacks are targeted at the human back end of MO rather than the software back end. The end goal doesn't appear to be gaining access to the site. $\endgroup$
    – François G. Dorais Mod
    Commented Nov 18, 2015 at 2:44
  • 3
    $\begingroup$ I'll somewhat echo @MadScientist and say that I would caution against anything that would require more effort on the moderator team's part in general than thwarting these "attacks". From my own experiences I know that moderating is occasionally much more taxing and time-consuming than it ought to be, and whack-a-mole is fun for only a very short time. But administering a policy against (some? new?) anonymous users may be even more time-consuming and taxing, with questionable efficacy. $\endgroup$
    – user642796
    Commented Nov 18, 2015 at 11:02
  • 10
    $\begingroup$ I support keeping the current policy regarding anonymity. When in comes to content I support the policy of deleting abusive contents/other content and be strict about it even in cases of doubt. When it comes to individuals, I support tolerance to a variety of ways in which people use MO and present themselves and anonymity is fully legitimate. $\endgroup$
    – Gil Kalai
    Commented Nov 20, 2015 at 6:28
  • 6
    $\begingroup$ "In view of recent attacks on MO..." Which attacks? $\endgroup$
    – Mahdi Majidi-Zolbanin
    Commented Nov 22, 2015 at 15:36
  • 10
    $\begingroup$ @MahdiMajidi-Zolbanin: We're mostly concerned with a resurgence of known troublemakers using elaborate false identity schemes to access the site. One such user has been known to impersonate others on other sites for defamation purposes. The situation seems to be calm now but there are still reasons to be cautious. $\endgroup$
    – François G. Dorais Mod
    Commented Nov 22, 2015 at 22:32
  • 4
    $\begingroup$ @FrançoisG.Dorais: According to what you know, has this user misbehaved to an extent where it may be a case for justice, or rather not so? $\endgroup$
    – Stefan Kohl Mod
    Commented Nov 22, 2015 at 22:46
  • 5
    $\begingroup$ @StefanKohl: Yes. One such early incident involved moderators and we decided not to pursue that course of action at the time. $\endgroup$
    – François G. Dorais Mod
    Commented Nov 22, 2015 at 23:03
  • 5
    $\begingroup$ @StefanKohl: Things are under control but not because of our change in policy. I would still advise caution with suspicious behavior. $\endgroup$
    – François G. Dorais Mod
    Commented Dec 9, 2015 at 17:44

6 Answers 6

Reset to default
71
$\begingroup$

There are two important reasons to allow anonymous users.

  1. There are countries/regions where western education, or women's education in particular is a dangerous issue. There are many more that monitor and censor their citizens' activity online. There are families that may be opposed to their children doing math. If you believe that MO has an educational mission, these users should be accommodated. This includes both pseudonymous and anonymous accounts and allowing connections via Tor.

  2. For early career mathematicians, MO can be pretty intimidating. We're all afraid of being wrong in public. Here's a fun fact: many women are more afraid to ask questions at conferences because if the question is bad, everyone will remember who asked it because you're the woman. Same story for minorities. This is a semi-professional setting, and that's terrifying. I was too afraid to post for years, and if I had to use my full name I still wouldn't.

$\endgroup$
16
  • 8
    $\begingroup$ For early career mathematicians, not to say grad students, MathOverflow can be a wonderful opportunity for increasing their networking skills. I always recommend the site, and always recommend to use real names. My experience with it had been great. Despite the bad things that happened, and would have been avoided if I had used a pseudonym. $\endgroup$
    – Asaf Karagila Mod
    Commented Nov 17, 2015 at 6:39
  • 7
    $\begingroup$ If real-name usernames were required on MO, I would not have joined as early in my career as I did, and it's likely I would never have joined at all. I suppose each user can weigh my reputation and decide whether that's a point in favor of real names or a strike against them (en.wikipedia.org/wiki/Book_of_the_Dead#/media/…). $\endgroup$
    – Vectornaut
    Commented Nov 17, 2015 at 6:47
  • 6
    $\begingroup$ But MO does not have an educational mission in any sense like implied in this answer. The cases listed are not generally applicable to those for whom MO is meant (this is the sort of thing MSE is good at). $\endgroup$
    – Tobias Kildetoft
    Commented Nov 17, 2015 at 8:19
  • 9
    $\begingroup$ @TobiasKildetoft point 2. seems fully applicable (and much less applicable to math.se), and point 1. too, depending on the interpretation (but I can see that the word "children" might have caused associations that lead to your comment). $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 9:34
  • 1
    $\begingroup$ I see these as reasons to allow people to post on MathOverflow using ( nonharmful ) user names. I do NOT see these as reasons to register with moderators using false or misleading credentials, or even no credentials. Can you give reasons for anonymous registration? :\ $\endgroup$
    – Gerhard Paseman
    Commented Nov 17, 2015 at 17:02
  • 1
    $\begingroup$ @GerhardPaseman Two things: first, the question among others asks "Should we require using real names?" (it says "using" not "providing"). While I agree it makes sense to keep these two separate. The question does not. I do not see why you would reproach this the answer. Second, depending on circumstances one may or may not be comfortable to provide such details even if it is claimed not to become public. I have confidence in the MO mods that they do not leak such info willingly or accidentally, and in SE that they set up the site so that it will not be comprised. $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 18:51
  • 2
    $\begingroup$ But, it is not unheard of that supposedly confidential account data gets compromised and made public. Now, I do agree that it is not all that likely, but still somebody might be worried about it (not I). What I agree about is that providing false information is an odd thing to do. Yet, I think you ask the wrong question: you ought to give reasons for personalized registration not the other way round. Especially in light of events that seem to show that this is not only pointless but rather causes problems the onus is more than ever on those that want it to present an argument. @GerhardPaseman $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 18:53
  • 1
    $\begingroup$ If I intend to reproach anything, it is the conflation (anon. username and anon. registration), not the question, answer, or their makers. I think the distinction should be emphasized. My major reason for personalized registration is accountability. I do not see either your reasons or EJI's as good reasons to "evade the moderators". Of course, there may be a good reason that benefits the forum and its community: let's see it. :(( $\endgroup$
    – Gerhard Paseman
    Commented Nov 17, 2015 at 19:01
  • 1
    $\begingroup$ @GerhardPaseman I agree that the distinction should be emphasized. And, I tried to give you some reasons why somebody might hesitate to give the data. Furthermore, what is "evade the moderators" supposed to mean exactly? What can they, what should they, do beyond the realms of the site? Specifically, what could the moderators do about your behavior that they cannot do about mine? Please be specific. (There may be situations where this could be relevant, but they should be extremely isolated and really even then it is truly not obvious what can be done.) $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 19:17
  • 4
    $\begingroup$ Your post is definitely me here. $\endgroup$
    – user82419
    Commented Nov 18, 2015 at 5:27
  • 3
    $\begingroup$ If you're afraid of being wrong in public, then mathematics is not for you. $\endgroup$
    – Christian Remling
    Commented Nov 19, 2015 at 1:16
  • 20
    $\begingroup$ @ChristianRemling Once I was afraid in the way you describe. Should I have been told mathematics was not for me? I like to think I have done tolerably well at it, eventually, since those days $\endgroup$
    – Yemon Choi
    Commented Nov 19, 2015 at 1:50
  • 3
    $\begingroup$ @YemonChoi: I had meant this as a rather light-hearted remark, along the lines of: if you do math for a living, you'll be wrong a lot, and some of the time in public, so why not get used to it now. (Also, by "you" you mean the OP? I was quoting from the OP.) $\endgroup$
    – Christian Remling
    Commented Nov 19, 2015 at 2:48
  • 1
    $\begingroup$ (1) I agree with EJI. Many papers in experimental psychology on this is. (2) While this is decisive, I do agree with C. Remling. Scientists often makes plenty of errors in public, should learn not to fear this, agreed. I've been wrong in public with utterly trivial (in hindsight) absent minded calculation errors. People point it out, I say thank you. Two questions on MO like that. I deleted those once I realized the mistake. If many nontrivial hypotheses are made, they can't all be right. Worst case all are wrong, best case all but one. That doesn't stop publishing. (1) is decisive tho here. $\endgroup$
    – Gottfried William
    Commented Nov 20, 2015 at 1:02
  • 6
    $\begingroup$ @Christian: Someone once jokingly said that being a mathematician is being wrong 99% of the time, and hoping to be remembered only for the other 1%. (Okay, cards on the table, that someone was me.) $\endgroup$
    – Asaf Karagila Mod
    Commented Nov 20, 2015 at 14:27
25
$\begingroup$

Adding to EJI's arguments in favor of anonymity (please upvote their post if you agree with my arguments):

[EDIT: These arguments are in favor of allowing anonymity by default; as far as I understand, this is what the OP invited us to reconsider. I do agree with banning anonymity for particular users that have a history of misconduct.]

  1. Some people work under overbroad confidentiality or intellectual-property restrictions, which make it risky for them to release any scientific output to the public without their employer's green light. (For an example, albeit rather old, see Student.) Even if only moderators were able to see proper names, these people might have reservations.

  2. When telling people to register anonymously, you are forcing them to place a nontrivial amount of trust in you, the MO community and the SE platform. Having spent 5 years on these, I can vouch for all three, although this has not always been this way (I remember not only me being mistrustful towards SE a while before the SEification of MO, back when Jeff Atwood was moderating). Should a newcomer who has just learnt about MO and sees a question they can answer have this trust right away? One of the things that made MO work so well is the ease of joining, and too high expectations on the user's confidence in MO will undermine this ease.

  3. There are people who do not want to be associated with controversy in discussions, even if they do not themselves actively take part in this controversy. You are forcing them to attach their names to discussions they cannot control. I remember one MO member threatening with leaving MO due to being accused of sexism in a comment thread on the grounds that a future employer might see the debate and file away their application as "possibly controversial" without bothering to check who was in the wrong. (I hope I did not misrepresent the argumentation.) While each of us has their own opinion on this sort of hiring practice, I think a number of people shares this kind of reservation, and MO will be off for the worse if they are kept out of participating. This is a mathematical knowledge-sharing site, not a test of courage, however easy this test might be.

  4. Having a policy in place is one thing, policing it is another. How are we going to police real names? Facebook has tried it, and the results have been a train wreck. Also, the canonical link to post at this point :)

$\endgroup$
10
  • 1
    $\begingroup$ I think you are missing the point that some people will be forced to use their real names. Not everyone at all. And the points you raise seem to me as an excellent reason for these people to be on their best behavior, rather than using anonymity as a shield for misbehaving. If you came to the site to troll, why would you care if you get suspended under a fake name? If you have to use your real name, then this might have real world consequences so you'd be less inclined towards misbehaving. $\endgroup$
    – Asaf Karagila Mod
    Commented Nov 19, 2015 at 3:41
  • 1
    $\begingroup$ @AsafKaragila: Ah! This wasn't all that obvious from the OP, which looked like an invitation to revise the anonymity-by-default policy. I am still unsure if it is me misreading the OP or you (or do you know its purpose from first hands?). To me it appears that the ban-on-anonymity-for-some-users has already been decided (which ban, in itself, does not break new ground: we have already banned some users completely, haven't we?), and this particular topic is bringing up the question of anonymity for all users. $\endgroup$
    – darij grinberg
    Commented Nov 19, 2015 at 3:44
  • $\begingroup$ Well, read this in light of meta.mathoverflow.net/questions/2573/… and meta.mathoverflow.net/questions/2575/… $\endgroup$
    – Asaf Karagila Mod
    Commented Nov 19, 2015 at 3:46
  • 7
    $\begingroup$ @AsafKaragila: I did, but as I said, to me it seems that this thread is about a wider-reaching real-name policy. "Should we ban any users that use anonymizing tools to hide their identity? Should we require using real names?" This is not about some known vandals. $\endgroup$
    – darij grinberg
    Commented Nov 19, 2015 at 3:49
  • $\begingroup$ I don't know... Maybe you're right. But I just don't think it's feasible at all to prohibit anonymity entirely. So I'm just not trying to read it that way. $\endgroup$
    – Asaf Karagila Mod
    Commented Nov 19, 2015 at 3:54
  • $\begingroup$ @Asaf: it may well be true that some users "[are] less inclined towards misbehaving" when "forced to use their real names." However, as said in my post, those that cause the real trouble often do so regardless the real name. (Mentally name the top-five SE users that caused you most annoyance. How many had used their real-name?) In the end this is also unsurprising, as those often are sure they are in the right anyway, and only due to some severe misfortune they happen to be around some of the very few that just can't see it; but the rest of the world knows or at least will know this. . $\endgroup$
    – user9072
    Commented Nov 19, 2015 at 13:21
  • $\begingroup$ @quid: Incidentally, the top three also used sockpuppet accounts (more than once each!), and each of them used these in a completely different way. But that's besides the point, I guess. :-) $\endgroup$
    – Asaf Karagila Mod
    Commented Nov 19, 2015 at 13:26
  • $\begingroup$ @Asaf: I take it this is a wordy way to say that at least the top three had used their real-name. :-) (But, yes, sock-puppets and ban evasion is a problem.) $\endgroup$
    – user9072
    Commented Nov 19, 2015 at 13:31
  • 1
    $\begingroup$ @quid: Although the three I recalled are not surprising for being there, and you're right. In their mind eyes, they had the right idea to begin with, and everyone disagreeing is wrong. But there were more than a few instances of pinpointed tackles with people, where they were obnoxious beyond words, and I'm afraid that those were mostly anonymous users (or at least the ones I can recall with ease). $\endgroup$
    – Asaf Karagila Mod
    Commented Nov 19, 2015 at 14:10
  • 3
    $\begingroup$ Points 3 and 5 are very good. Too bad answers can't be joined into one with two authors. (This plus EJI's as top post.) In last three years, I remember somebody I knew fired for publishing without permission (work unrelated to their job entirely, because they were in industry and the company had a broad permission for publishing anything policy. (Because each worker is affiliated with the employing entity.) $\endgroup$
    – Gottfried William
    Commented Nov 20, 2015 at 1:26
24
$\begingroup$

What anonymity policy should MO have?

You say "As an internal moderation policy, we've always had less tolerance toward anonymous users but we never set any official rules or guidelines." This was repeatedly said, and in my understanding this is the policy. It is also in some old FAQs, IIRC. It seems it worked reasonably well.

In view of a comment let me add that I am not sure it makes sense to write down some suspension policy. Cases are individual and need to be handled on an individual basis. To fix some rules in abstract seems designed to lead to problems, and endless discussions. It seems reasonable to me not to be very restrained when facing users that seem to try to use the site in bad faith. Whether they do so under the auto-generated name, a pseudonym, their real-name; or a "real name."

The comment of Mad Scientist is in my opinion spot-on. Mainly, moderator action should be taken against behavior, not based on some secondary criteria. Indeed, doing the latter creates the separate issue that users try to work around it, and in doing so do things that might well be considered quite more problematic than anonymously asking a question on this site (even when in doing so they evade a ban).

Should we ban any users that use anonymizing tools to hide their identity?

Assuming you mean TOR and alike. Some users, not me, might have a need to use such technology that has nothing to do with trying to misbehave on MO. See for example this request of a real-name user. I do not think to just ban this on principle is a good idea. It could be consider as a factor in moderation decisions though.

Should we require using real names?

Correct me if I am wrong but as far as I know and recall several (not to say most) of the bigger problems were actually caused by users under their real-name (or at least under what I took to be their real name, see below) or latter additional accounts that were, to those in the know, linkable to real-names. Thus, I do not quite understand what should be achieved by this.

Indeed, one might go as far as speculating that the real-names in the end had a negative effect. Users had to dig themselves in rather than to have the option to just walk away and perhaps come back in good. Or, they pulled off stunts to get around the real-names encouragement policy.

On a practical note: there are masses of users in the network that might create an account with one click. Most of them will not be aware of such a policy. Would you intend to follow up on each such instance or delete on-sight? This will cause a lot of work and friction. Or, there would be a need for a technical solution.

Basically, I think one would have to change the sign-up drastically. However, then real names as display names would (still) not really be needed either. In fact it would be less needed then.

Should we maintain the status quo?

Yes, mainly. What annoys me personally are "fake real-names." Users should either use their real-name or something easily identifiable as not a real-name. If such "fake real names" would be discouraged or essentially banned it would have my support. (And I'd yet have to hear an argument in favor of them. There existence might also be encourage by the existing real names policy, one more point against it.) I know there might be some corner-case, where this might cause some issue, but by and large this should be reasonable to maintain. One might also make exceptions for common given names. But something that quite definitely looks designed to be read as a given-name name combo should be real or better avoided.

Are there other things MO could do to encourage civil and responsible behavior for all users?

You could moderate more tightly. I do not mean bans or moderator messages. Just delete comments that seem a bit rude or even just snarky (there is no need to make much ado about it, they will just be gone). On other SE sites moderation is a bit stricter. They have no real-names policy but are rather more civil (well some of them, some of the time, and I do not mean the one I moderate, which in any case is too small to be representative for anything).

Finally, I think one should resist the temptation to try to install policies one cannot enforce. It can back-fire. Some months ago (maybe longer) a user in a meta-post was quite explicit about this. They got banned permanently, and they try to evade the ban, and basically they have nothing to lose. While I do support that ban (even actively at times) and that user did things completely out of line, it still shows a problem with permanent bans.

Personally, I feel that at times moderation on this site is too emotional. At first, little is done since we are all a big community and one does not want to be seen as being oppressive or censoring. Then, things go too far, and many lose their patience and relatively massive actions are taken.

I believe sometimes it would be good issues would be escalated earlier but slower. The site has quite clear rules regarding content and conduct. They should be enforced on a daily basis, even if at times only by a stern comment.

$\endgroup$
1
  • 4
    $\begingroup$ "I think one should resist the temptation to try to install policies one cannot enforce. It can back-fire." Agreed 100%. This is the decisive argument for this thing. $\endgroup$
    – Gottfried William
    Commented Nov 20, 2015 at 0:46
13
$\begingroup$

I think whether or which changes should be made to the anonymity policy of MO depends crucially on the nature of the recent "attacks" on MO, and without more information on them being disclosed, a sensible discussion of the issue is hardly possible.

That said, on the one hand I think that EJI has raised good arguments in favor of allowing anonymous contributions, but depending on the nature of the "attacks" it is quite possible that it is necessary to become more restrictive and require using real names and e.g. establish an arXiv-like endorsement system.

$\endgroup$
3
  • 2
    $\begingroup$ The question of restricted sign-up and real-names need to be used as display-names are quite orthogonal. On the "attacks." I cannot know if this is what is referred to but what I noticed happening was that numerous "new" users asked question to create what can be seen as some kind of flooding-attack (it was not "spam" only substandard question, but not very bad either). It seems to me requiring "real names" would have done nothing to prevent this. Restricted sign-up would, but this would need to be implemented, and the question is if one really wants to go that far. $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 11:32
  • 1
    $\begingroup$ @quid: At least in the last 10 days, from the post numbers I see no indication of a large number of deleted questions, and also there don't seem to be unusually many downvoted or closed questions such as to speak of an "attack". Or what do you refer to? -- Given what the moderators say, I would have thought it's something notably more serious. $\endgroup$
    – Stefan Kohl Mod
    Commented Nov 17, 2015 at 11:57
  • 4
    $\begingroup$ As I said it was "only substandard question[s], but not very bad either." By now several of them are answered or up-voted. Further, the user names changed and meanwhile look like real names, further illustrating my point. When you think about potential implications, this could become quite serious if it is malicious. $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 12:09
-3
$\begingroup$

I think the real issue is trying to game the users and community toward proper behaviour, and that anonymity of user name is not the issue, it's attempted anonymity behind the email address/verification system/way to hold a person controlling user accounts to be responsible for their actions that is the issue. Perhaps instead of anonymity, we should be using the term evasion.

Edit 2015.11.17 in response to comments of quid and of mine. The discussion of whether to have anonymous usernames is not new, and some arguments for them are repeated here. The anonymity policy should also address the issue of registering to use the site. Below I advocate a policy of non-anonymous registration, as well as adjust how problematic registrants are treated. I do not say to remove bans; that is a part of the administrators toolkit. I do say that community treatment as well as moderator treatment of such registrants should change.

To address some of quid's concerns raised in comments: I recommend non-anonymous registration to increase accountability. The vitally private registration information (real name, real contact, real endorsers, other real info) does not need to be made available to anyone but the top level adminstrators, and should not be needed except as used to hold registrants accountable. This ideally is limited to behaviour moderation and control on the forum network, but can extend to real-life reputation, as such information could be used in legal or civil or public actions against the registrant. Play nice, don't screw up too much, and noone ever needs to look at the record. Do something for which you need to answer to the moderators and or forum community, then the guardians have something to use. When I ran an internet access company, we needed and guarded such records for adminstration and accountability purposes. If the dogs are attacking, the Internet dogcatchers need to know where the dogs are. You want examples quid? Ask the moderators; I don't have the information and prefer not to speculate in this thread. Also, I don't need to know your real name or have the moderators treat you and me with seriously different policies. I need, you need, the community and the health of this forum need the moderators to know how to hold me, you, and the community members accountable. Anonymous registration works against this. End Edit 2015.11.17

It may be time for a two- or multi-track system, which is a little more elaborate than suspend/unsuspend/ban. For simplicity, I describe a two-track system and let the social engineers generalize it. Suppose there is an identification system (not necessarily Open-Id, say AMS membership for example, or an endorsement from an existing forum member in good standing) which allows one to register for participation in the forum. Then such a validated person is on the first track, can open multiple accounts, and participate with real or fake names as wanted, within restrictions to reduce harm that might be caused.

There is no need for a second track until misbehaviour is identified, warned, and gone uncorrected. Then the second track kicks in, and the appearance and other properties of the accounts change. Primarily, reputation gets harder to earn for these accounts ( and may be reset to zero ), and all posts from this person appear in a yellow background. Only model behaviour and exceptional circumstances can take a person off the second track back on to the first, and the system is designed so as to discourage posts from this person while on the second track. This is different from banning because every contribution the person chooses to make is a signal to themselves and the community that they are on the second track, and are working their way back to the first track; banning or suspending just prevents them from making positive or negative contributions.

It may be time to abandon the ideal of "let everybody have a chance to post", and move towards allowing only those who have shown some ability to mesh with the community to allow to post on MathOverflow. Having them register with and participate well on other forums first may allow them easier entry into a specialized forum like the one this community has built.

(I am suspending signatures on this thread until such time as a relaxed and welcoming atmosphere can be restored. GRP)

$\endgroup$
2
  • $\begingroup$ Thanks for the details. As I said "There may be situations where this could be relevant, but they should be extremely isolated and really even then it is truly not obvious what can be done." Your description confirms this. In this regard I'd like to point out that the formulation "Do something for which you need to answer to the moderators and or forum community, then[...]" is quite misleading as the moderators can in fact contact any user directly. $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 22:56
  • $\begingroup$ It is more "Do something so that the mods see fit to send the police to your home" where your scenario applies. It could happen, but I am not sure it makes sense to design the sign-up process around this idea. But, to find a point of agreement, to insist on verifiable and verified identification on sign-up would have some actual effect. (Though I continue to think the emphasis is completely misplaced.) $\endgroup$
    – user9072
    Commented Nov 17, 2015 at 22:56
-3
$\begingroup$

One possibility is to require of everybody to disclose real identity to the moderation team, with an option to disclose or not disclose it to everybody; rigorous formulation of rules about which kind of behaviour may then imply that the identity must become public is not necessary in my opinion.

(The comment by quid indicates I've been too fuzzy here so I'll try to detalize)

If I understand correctly the idea is that if a real person behind some sort of abuse might become publicly known this would likely reduce motivation to commit that abuse. In this case it is even more efficient not to reveal precise rules describing in which cases such public disclosure will take place: the rest of the users will not be concerned anyway. All that is needed is for the moderating team to be able to provide information about real identity of a user in case of any inquiry or even without it. When exactly to do this I think it is completely safe and even more efficient to leave them to decide in each case individually.

(Hope this was not even more fuzzy...)

If this requires lots of technical work, and not only on MO but on the whole SE network, still I believe it is worth it, because of several serious reasons already well explained in answers here.

In any case I think discriminating among "suspicious" and "nice" users by requiring different things from them is a very harmful idea.

$\endgroup$
18
  • $\begingroup$ Could you please explain more precisely what you mean by: "then imply that the identity must become public." What exactly do you envision to happen? $\endgroup$
    – user9072
    Commented Nov 24, 2015 at 10:38
  • $\begingroup$ I've tried to explain. I believe the only thing that matters here is to know that it is known to moderators who committed this or that bad thing. I believe this will have practically the same effect as if all users would be immediately identifiable with real life persons. $\endgroup$
    – მამუკა ჯიბლაძე
    Commented Nov 24, 2015 at 13:11
  • 2
    $\begingroup$ Thanks for the clarification. To be able to do this would need a (quite drastic change) of the the privacy policy of the site. I think basically a user would have to agree in advance that the moderators can make public their private data (or specified parts thereof) for any reason. Of course, this the would have to be agreed to by all existing users, too. $\endgroup$
    – user9072
    Commented Nov 24, 2015 at 13:27
  • $\begingroup$ @quid Yes. But I believe it is worth doing in the existing circumstances. $\endgroup$
    – მამუკა ჯიბლაძე
    Commented Nov 24, 2015 at 13:40
  • $\begingroup$ How would what you propose help to solve problems in "the existing circumstances" (and what are these circumstances)? In particular, why should somebody "using elaborate false identity schemes to access the site" care about the false identity being made public? $\endgroup$
    – user9072
    Commented Nov 24, 2015 at 13:48
  • $\begingroup$ @quid By "existing circumstances" I meant only that part which is visible to me: that quite many harmless users seem to care much about retaining right to use pseudonyms or otherwise have their real identity undisclosed to the public. $\endgroup$
    – მამუკა ჯიბლაძე
    Commented Nov 24, 2015 at 13:52
  • $\begingroup$ As for your second question - it remains valid wrt any other possible proposed solution I think. If moderators cannot know for sure whether the announced personality is genuine or fake, nobody else can too. $\endgroup$
    – მამუკა ჯიბლაძე
    Commented Nov 24, 2015 at 13:54
  • 1
    $\begingroup$ It appears we are talking at cross purposes. On the one hand you say many harmless users care about retaining their privacy, on the other hand you propose a scheme that means these users must waive all rights to this privacy before being allowed to use this site. I think I have to misunderstand something there. On the "second question." I do not see how what I proposed is affected by this problem, in a nut-shell it is: an account that behaves reasonable can stay, an account that misbehaves is shown the door. $\endgroup$
    – user9072
    Commented Nov 24, 2015 at 14:02
  • $\begingroup$ @quid Simply there is big difference between facing immediate public recognition and having to believe in moderators' good will. Concerning your proposal - I believe it will only encourage more and more aggressive repeated attempts. $\endgroup$
    – მამუკა ჯიბლაძე
    Commented Nov 24, 2015 at 14:04
  • $\begingroup$ (But I admit it is a counterexample to what I said (that that question of yours applies to any proposed solution)) $\endgroup$
    – მამუკა ჯიბლაძე
    Commented Nov 24, 2015 at 14:09
  • $\begingroup$ I do not agree with your penultimate comment. But it might not be feasible to discuss this here to the extent needed to find some common ground. Perhaps let us agree to disagree on that. $\endgroup$
    – user9072
    Commented Nov 24, 2015 at 14:13
  • $\begingroup$ Sure thing :D Actually I am neither competent nor experienced with such matters, so... $\endgroup$
    – მამუკა ჯიბლაძე
    Commented Nov 24, 2015 at 14:17
  • $\begingroup$ @quid Sorry only one more thing - do you disagree with both parts of it? $\endgroup$
    – მამუკა ჯიბლაძე
    Commented Nov 24, 2015 at 14:34
  • 5
    $\begingroup$ I think the problem with your scheme (giving the moderator team your real identity) is that it means not just trusting the current moderator team at this point in time, but trusting them for an indefinite time, trusting all possible future moderators, and the organisation behind stackexchange and mathoverflow for all time. If I were a pseudonymous user for any of the reasons other people stated in this thread here, I would not agree to this, as there are too many uncertainties. $\endgroup$
    – Jan Jitse Venselaar
    Commented Nov 26, 2015 at 9:30
  • 6
    $\begingroup$ @მამუკაჯიბლაძე What I was trying to say is: you're not only trusting the current moderator team, but all possible future moderator teams as well. Communities change, and so will the moderator team. I don't think it is fair to promise people that their names will be kept private, when in fact it is impossible to guarantee so (even leaving aside the issue of hacks). I think the issue of trust will also vary from person to person. I trust the moderators (in fact, the internet) with my name, but other people might have had worse experiences, and will want to keep it from people they don't know. $\endgroup$
    – Jan Jitse Venselaar
    Commented Nov 26, 2015 at 16:26

You must log in to answer this question.

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