Case Opened on 18:55, 29 October 2007 (UTC)

Case Closed on 12:06, 18 November 2007 (UTC)

Please do not edit this page directly unless you wish to become a participant in this case. Only add a statement here after the case has begun if you are named as a party; otherwise, your statement may be placed on the talk page, and will be read in full. Evidence, no matter who can provide it, is very welcome at /Evidence. Evidence is more useful than comments.

Arbitrators, the parties, and other editors may suggest proposed principles, findings, and remedies at /Workshop. That page may also be used for general comments on the evidence. Arbitrators will then vote on a final decision in the case at /Proposed decision.

Once the case is closed, editors may add to the #Log of blocks and bans as needed, but closed cases should not be edited otherwise. Please raise any questions at Wikipedia:Requests for arbitration#Requests for clarification.

• Sadi Carnot (talk · contribs · deleted contribs · logs · filter log · block user · block log)
• Coren (talk · contribs · deleted contribs · logs · filter log · block user · block log)
• Kww (talk · contribs · deleted contribs · logs · filter log · block user · block log)
• Jehochman (talk · contribs · blocks · protections · deletions · page moves · rights · RfA)
• Physchim62 (talk · contribs · blocks · protections · deletions · page moves · rights · RfA)
• Sarah (talk · contribs · blocks · protections · deletions · page moves · rights · RfA)
• DragonflySixtyseven (talk · contribs · blocks · protections · deletions · page moves · rights · RfA)

User:Coren filed a report at Wikipedia:Administrators' noticeboard/Incidents stating, "After the closure of Wikipedia:Articles for deletion/Georgi Gladyshev...I've spend some amount of time looking through the contributions of Sadi Carnot. What I see there is a large, elaborate a subtle walled garden of pseudoscience— probably for the purpose of hawking his books (or simple self agrandizement)."[1] I recognized Sadi's unique name because I had mediated a Wikipedia:Mediation Cabal case between Sadi Carnot and Hkhenson. (I was on good terms with both.) I added my comment to the thread. Then User:SandyGeorgia, a well-respected editor, stepped forward with additional evidence of disruption by Sadi Carnot. Upon further investigation we determined that Coren's summary was most likely true, and that there were at least 132 of Sadi Carnot's spam links woven throughout a network of articles, all pointing back to his own website. More spam links were later found, and three sites affiliated with Sadi Carnot have been added to the MediaWiki blacklist.

We also discovered that Sadi Carnot had previously used an account User:Wavesmikey. This account was apparently abandoned after it was exposed for spamming and pushing fringe theories. Over the past two years Sadi Carnot has undertaken a campaign to spam his personal website and push his own fringe theories across a wide variety of articles. In the process, he has subtly vandalized many articles by inserting pseudoscience and by misrepresenting sources. Numerous editors attempted to stop him along the way, to no avail. His strategy was to retreat when confronted, and move on to other articles. In light of this evidence, User:Kww proposed a community ban.

In my judgment, Sadi Carnot knew that what he was doing was wrong because he used deception. He had also cited the conflict of interest guideline against User:Hkhenson during the Mediation Cabal case. Subtle vandalism is more dangerous than page blanking or curse words because the average editor does not realize they are reading falsified information. Considering the risk of further damage, I placed an indefinite block on Sadi Carnot, and suggested that this block could only be lifted if he accepted responsibility for what he had done, and made arrangements to be mentored by an experienced Wikipedian. I announced this block at ANI, and other editors notified the relevant WikiProjects, which is how I believe Physchim62 learned about this matter.

User:Physchim62 proceeded to revert my block without prior discussion, in spite of the fact that Sadi Carnot had not even requested to be unblocked, and that no arrangements had been made to monitor his editing. Subsequently, a large number of editors endorsed the community ban proposal, and User:Sarah reinstated the block. Despite the overwhelming consensus that Sadi Carnot should be banned, Physchim62 contacted DragonflySixtyseven via IRC, and IRC activities resulted in DragonflySixtyseven unblocking Sadi Carnot a second time, also without prior consultation with the blocking admin, or those of us who had spent many hours investigating Sadi Carnot's contributions.

Members of the arbitration committee, we have an "unusually divisive dispute among administrators." To prevent further conflict, I request that you review this case and provide guidance how the community ban process should work so we can avoid situations where a small number of good faith administrators frustrate the community. The community banning process does not give each of 1400 administrators veto power. Per common sense, that would be unworkable. What is the point of having a consensus discussion if a single administrator can be convinced by a troll to revert a block that has been endorsed by the other 1399 administrators? We must not allow disruptive editors to continue gaming us and preying on our generous assumption of good faith. - Jehochman ^Talk 00:37, 23 October 2007 (UTC)[reply]

"Warning is not a prerequisite for blocking (particularly with respect to blocks for protection) but administrators should generally ensure that users are aware of policies, and give them reasonable opportunity to adjust their behaviour accordingly, before blocking."

— Wikipedia:Blocking policy

Sadi Carnot was responsible for long term spamming, long term fringe theory pushing, and long term COI editing. He was well aware of policies, and he had many, many, many opportunities to adjust his behavior.

"If Sadi Carnot wishes to continue editing, he can ask to be unblocked, and we can discuss the conditions on which that will occur. An indefinite block doesn't mean "forever." It means, "until the problem is resolved." I personally wouldn't unblock him until he recognizes that what he's been doing isn't compatible with Wikipedia's purpose, and he undertakes not to edit the articles or subject areas where problems have occurred in the past. Additionally, he should join WP:ADOPT to be paired with an experienced editor who will monitor and assist his editing to make sure there are no relapses. Bduke, I think Sadi Carnot may become a good faith editor, but right now he doesn't understand how to edit Wikipedia and he's causing tremendous damage that involves many articles. My block is designed to prevent further harm until we can come up with a better arrangement."

— Jehochman

"I think the use of multiple accounts is very problematic. If you look at the sheer volume of self-promotional links and POV pushing, this looks like a determined COI editor who does a few good edits to establish cover. Of course, as I said above, if the editor is willing to admit mistakes and agree on editing restrictions and mentorship to avoid further problems, I am open to him returning. However, I think it would be a serious mistake to let him edit again before we have those assurances."

— Jehochman

I made these comments before Physchim62 unblocked. Those who cry "witch hunt" need to review the facts. - Jehochman ^Talk 23:35, 24 October 2007 (UTC)[reply]

Possible mootness

(Update) Physchim62 and I have come to an agreement that he will mentor Sadi Carnot should he return, which seems very unlikely at this time. Sadi has received a final warning and will instantly be blocked if problems resume. Sadi's websites have been blacklisted, so there is no way to spam them. I believe this controversy arose primarily because of serious defects in the Wikipedia:Banning policy. Rather than ask Arbcom to rewrite the policy or even recommend policy, I have opened a policy RFC at WT:BAN#Request_for_comments:_Community_bans. Everyone is invited to help fix the policy. That's where our energies should go. If anyone wishes to pursue this case, I won't have much to say. - Jehochman ^Talk 03:36, 26 October 2007 (UTC)[reply]

In response to Newyorkbrad's comments, I've been discussing, and have updated WP:BAN to reflect the idea that administrators should discuss unblocking before doing it, in order to prevent wheel warring. If Physchim62 and DragonSixtyseven agree that they should have discussed unblocking and worked toward a consensus rather than reverting administrative actions unilaterally, I think there is no longer anything to discuss. - Jehochman ^Talk 15:01, 27 October 2007 (UTC)[reply]

As the threat of arbitration was lifting, Physchim62 posted two troubling comments at WT:BAN that he still supports reverting admin actions without discussion. He also repeats the allegation of "witch hunt", which I consider to be an offensive breach of assume good faith, and civility because I have asked him several times to stop saying that. I'm not sure what to do about this. I've bent over backwards to try to resolve this dispute, yet he keeps throwing darts. What are we going to do? - Jehochman ^Talk 14:39, 28 October 2007 (UTC)[reply]

This big issue here is whether we already have, or should implement, 0RR for admin actions. I think that admin actions should not be reverted without discussion and agreement. When there's no agreement, there's arbcom. As the recent Jimbo-Miltopia-Zscout370 drama has shown us, Bad Things™ happen when admins revert each other, and the community is hopelessly confused. Perhaps this is the ideal test case to establish a lasting precident. - Jehochman ^Talk 20:56, 28 October 2007 (UTC)[reply]

I became aware of the problem with Sadi Carnot during the AFD discussion on Human chemistry. After reviewing his contributions, the pattern became very clear: distortion of his sources, used to prop up a pseudoscience agenda. Most of the edits I checked that did not directly support his particular form of nonsense were incorrect or strangely biased. I am left with two choices: he is either a fool or a con artist. If he is a fool, he can be morally excused, but Wikipedia still needs protected from him. If he is a con artist, he doesn't even get to be morally excused, and Wikipedia still needs to be protected from him. My personal belief is that he is a con artist, as he seems to be too literate to be a fool. Either way, it will take a virtual Wikiproject simply to undo the damage he has already inflicted, and there is no justification for allowing him to wreak more. It should also be noted that I grew sufficiently suspicious of User:Physchim62 to request a checkuser on him to ensure that he is not a second sockpuppet of Sadi Carnot.Kww 01:17, 23 October 2007 (UTC)[reply]

Late addition in response to some of the other comments. First, I am sorry that I speak bluntly. Coren's mens rea is simply Latin for the difference between a fool and a con artist, and it is apparent that he wrestled with the same question and came to an identical conclusion. Second, there are discussions of a Jekyll/Hyde aspect to the case. I beg to differ on this point ... I have not found a non-trivial valid edit. I welcome correction on this point.Kww 13:48, 23 October 2007 (UTC)[reply]

One more comment: I think the person to be assigned the role of mentor of Sadi Carnot (if that is the output of this), should not be either of the admins that lifted the block. Since they apparently don't agree that his current behaviour is unacceptable, I'm not sure that they will police him with sufficient vigilance.Kww 18:13, 23 October 2007 (UTC)[reply]

I think that letting such an acrimonious issue drop without arbitration would be a shame. I've registered objections to the proposed solution (that Physchim62 should not be permitted to mentor Sadi), which Jehochman has dismissed as a "detail." We need a definite, concrete solution which allows us to prevent Wikipedia from being corrupted by pseudoscientists and frauds. We need to make it clear to the admins that unblock pseudoscientists and frauds that such actions are unacceptable. I don't think the proposed solution accomplishes either of those goals.Kww 13:37, 28 October 2007 (UTC)[reply]

While I cannot agree with Kww's exact words, which are stronger than I would use, I nonetheless agree with the essence of what he is saying.

I brought Sadi Carnot to AN/I's attention after my involvement with the Georgi Gladyshev AfD, where I began to figure out that he has weaved a complex and subtle walled garden of fringe original research. After collective the digging on AN/I, I saw how deep the rabbit hole went— dozen of scientific articles subtly twisted with the ancient (18th and 19th century) science he appears to be enamored with, and a pattern of shoring up his own original research with references to articles he himself twisted to fit. The end result is the rather successful insertion of his pet theories in a way that is sufficiently self-consistent and apparently supported so that anyone but topic experts is likely to be fooled by the disinformation. Other contributors to the AN/I thread have explained his lack of comprehension of basic subject matter in the articles he edits vehemently.

In addition, SC has a demonstrated pattern of using sources that either ultimately point back to his own pseudoscientific web site (now blacklisted on meta), or that are blatantly misrepresented as supportive of his statements when in fact they are either unrelated or state the opposite of his claims.

Sadi Carnot has clearly shown mens rea by the nature of his sources, and by his repeated pattern of leaving when challenged only to return discreetly some time later.

Some editors have stated they feel that the number of his good edits make a ban too strong a measure. However, I feel that the skill of his alterations in articles he has edited so that expert attention is needed to ascertain their reasonableness makes all his edits suspect. I remain unconvinced that any of his edits should be taken at face value, especially since those would have come after the same pattern of OR insertion with his previous identity.

While I would not have supported a ban of an editor without warning, and originally felt that move might have been a bit heavy-handed; the extent and breadth of the damage he has caused to the encyclopedia that has been revealed in the AN/I discussion, and the fact that this damage is deliberate, skilfull (thus harder to repair), and has been going on for over two years make me agree that this is the only reasonable course of action.

SC would, of course, be welcome to mend his way and rejoin the community by appealing the ban in the normal manner. In the meantime, this would allow administrators to ban and revert recurrence of this editor on sight so that he is not allowed to repeat his previous move of switching identities and starting over.

— Coren ^(talk) 02:25, 23 October 2007 (UTC)[reply]

Further note: I want to make clear that I support this arbitration request to help clarify what is proper to do in cases where there is (in my opinion, anyways) strong evidence of long time, willful, tampering with the encyclopedia. I've begun digging edit-by-edit in the past history and amassing evidence for that.

I would not support and arbitration regarding the conducts of any administrator in this case. I think it's painfully clear that everyone here has acted in good faith but that there is a fundamental disagreement about Sadi Carnot's action in particular as well as how to react to long-term damage in general. Apparently, the very nature of this case is such that it hits strings very close to some administrators' hearts and we need to proceed carefully lest tempers flare. — Coren ^(talk) 15:50, 24 October 2007 (UTC)[reply]

I became involved in this dispute after reviewing the discussion on ANI requesting a community ban of User:Sadi Carnot. User:Jehochman had implemented an indefinite block which was overturned by User:Physchim62 without discussion and without any pending unblock request or undertaking from Sadi Carnot to cease spamming. I eventually reblocked, believing that there was a demonstrably clear consensus to block: at the time of my block many people had commented at ANI and while there were a small number of people who did not support or oppose the block and proposed ban, other than the sock of a banned editor, Physchim62 was the only editor who opposed.(This was the state of the discussion when I blocked) I feel that Physchim62 acted arbitrarily and outside community consensus and that this dispute reflects a profound divergence in the implementation of WP:BLOCK and WP:BAN and the perceived validity of community-based sanctions. I think the original block should have remained in place while the community continued discussion (after all, an indefinite block does not mean "forever", only that no time has yet been set) and pending a response and explanation from User:Sadi Carnot, which he could have made on his talk page. I am disappointed to see that the block has been overturned again, apparently in response to IRC activities. And I note that while User:DragonflySixtyseven did post on ANI saying that he was unblocking Sadi Carnot, he didn't bother to inform me that he had overturned my block or attempt to discuss the block with me personally, as is customary when overturning another administrator's actions. I object to the way unblocking this user has been treated by two administrators as a matter of urgency, particularly when the blocked user has not edited since 11 October. I request the committee accept this case to examine the actions of all parties and to confirm the validity of community-based sanctions. Sarah 02:57, 23 October 2007 (UTC)[reply]

Sorry about that - I've been having computer problems, and then I spent several hours dealing with a paying client; otherwise I certainly would have notified Sarah. Again - I apologize. To clarify: Physchim62 did not ask me to unblock Sadi Carnot. I made the decision independently - and WP:BOLDly - after a heated discussion in PM (not on a channel) on IRC, which resulted from my disappointment in PC62's mention that he would be giving up his admin bit as a direct result of the Sadi Carnot mess. He specifically said that he did not want to wheel war over this (since he considers many of SC's contributions to be unmitigated garbage), but that he was considering taking this to arbcom on principle. I consider this entire mess to be a waste of Arbcom's time and resources.

This is entirely a point of principle. Frankly, I don't think SC is much of a useful contributor - but he's clearly not totally useless. However, that's not really relevant since he has apparently left Wikipedia. The unblock was not for SC's benefit, it was for PC62's benefit - specifically so that he not give up his bit. I got him to agree to mentor SC, and to agree to block SC himself' should SC start acting up again (assuming, of course, that SC does actually return to the project, which is not a guarantee).

I also left a very very very stern message for SC. That is his warning. He has now been warned. If he (returns and resumes editing and) fucks up again, then we ban him. Simple, yes?

I decided that the potential loss of PC62 as an admin outweighed the risks of having SC unblocked (especially since SC has been warned, told that it's his only warning, given a mentor, and effectively placed on Topic Probation).

And you have no idea how tired I am. It's past one AM here. DS 05:06, 23 October 2007 (UTC)[reply]

I am delighted that so many editors have chose to comment on this case. I can only hope that all participants have read The Crucible by Arthur Miller, a great description of how mass hysteria can develop among otherwise reasonable people. To say that someone is a victim of mass hysteria is to say that they are a victim, it does not imply a value judgement.

I cannot answer all the issues raised by my colleagues without breaching the brevity guidelines for this page. I would just like to underline three points.

1. The disputed block is an indefinite block without warning of an editor who has contributed for nearly two years, who had a clean block record and 8537 contributions.
2. The reasons given for the original block are demonstrably false. Jehochman did not make the necessary verifications before taking such a drastic action. As has been admitted, there was no urgency to block as the account is currently inactive.
3. Banning policy, as it stands, allows any admin to veto a Community Ban. As such, there was no reason for Sarah to reinstate the groundless block. Her comments indicate that she had not correctly read the ANI discussion, let alone the underlying evidence. The contention that the few people who commented on ANI somehow represent the "community" is without basis. I might add that the "consensus" on ANI was nowhere near as clear as has been suggested.

It is my honest and deeply held opinion that Sadi Carnot is being persecuted not for what he did but for what he believes. I certainly don't subscribe to his beliefs, but he has an inalienable right to hold them. You will never be able to count on me to form part of a Thought Police.

Allow me end by apologising for the harsh terms which I use. I would not use them but for the seriousness of the situation. I've already had to warn one user that his actions relating to this dispute amount to disruption of Wikipedia: maybe more cases will come to light. I really believe that you're all going mad, and that it's not your fault. Physchim62 (talk) 17:02, 23 October 2007 (UTC)[reply]

To reply to those users who have said that my actions and those of Dragonfly6-7 are "frustrating the community", I would remind them that the community has abolished the Community sanctions noticeboard, in a MfD debate which attracted far more contributions than the discussion on Sadi Carnot's conduct. I cannot believe that it was the intention of the community that the failings of CSN should be repeated on ANI, and it is surely "frustrating the community" to act as if that were the case. Physchim62 (talk) 12:00, 24 October 2007 (UTC)[reply]

To give participants some idea of what is the Community, a few statistics (thanks to uninvolved Gmaxwell):

• one million unique IPs edited English Wikipedia in September 2007
• our current best estimate for the number of readers is around 200 million (the exact figure depends on how you do the estimate, obviously)
• Wikimedia Foundation projects currently receive 7 billion pages views per month.

The Community is not a dozen users on ANI, at least two of whom were acting in astounding (and in my eyes culpable) bad faith, something which the blocking admins should have detected and taken into account. Physchim62 (talk) 12:28, 25 October 2007 (UTC)[reply]

On FloNight's comments

I find FloNight's comments on accepting this case quite frankly offensive. It is common knowledge that the present ArbCom cannot cope with its current case load. This is why administrators work day and night to try to avoid bringing cases here.

ArbCom cases are WikiTorture for all involved. I have always considered this to be a Good Thing (I have a sneaking suspicion that it's deliberate) as it encourages users to take the other forms of DR seriously. If ArbCom wishes more cases to come here, the ball is in its court, but it will have to resolve those cases more quickly, not, for example, leaving cases langering in the voting stage for over a month as has been happening recently. It should remember that taking more cases is inconsistent with encouraging the use of other forms of DR.

FloNight's comments are also remarkably prejudicial. Does she think that administrators are some kind of playground bully, that we go around picking fights for the fun of it and, when we can't find any other poor user to pick on, we fight among ourselves? What ever my criticisms (and I have many) of the actions of Jehochman and Sarah, I do think they were trying to resolve a dispute, not to create one.

If FloNight feels unable to withdraw her comments, she should recuse herself from this case. Physchim62 (talk) 12:00, 24 October 2007 (UTC)[reply]

The "solution" to which Newyorkbrad refers to below can be found here. It was intended as an interim solution until such time this case could be resolved by other means. I cannot see how admins should be lectured on why they should bring cases to ArbCom, only for ArbCom not to hear those cases because the same admins have agreed to wait for more discussion (ie, that the evidence is presented in full here). It has been pointed out by Newyorkbrad himself that current banning policy is inconsistent, indeed that ArbCom policy on this matter is itself inconsistent. I would remind members of this Committee that their decisions are enforced by administrator and community action: if they lose the confidence either of administrators (as a group) or of the Community as a whole they will have succeded in destroying a useful tool on Wikipedia. Physchim62 (talk) 13:25, 28 October 2007 (UTC)[reply]

Allow me to suggest two simple questions for ArbCom to decide:

1. Was Jehochman's block correct?
2. Was my unblock correct?

If this committee could bring itself to consider these two controversial administrative actions, it would already have 90% of the reply to the rest of the points raised by users in this request. Physchim62 (talk) 18:02, 28 October 2007 (UTC)[reply]

• Accept. Kirill 00:46, 23 October 2007 (UTC) Reject, per Newyorkbrad. Kirill 15:43, 27 October 2007 (UTC) And accept again, unfortunately. Kirill 21:00, 28 October 2007 (UTC)[reply]
• Accept. Blnguyen (bananabucket) 01:55, 23 October 2007 (UTC)[reply]
• Accept. At first glance the problem seems to be administrators unblocking an editor, that one or more administrators believe is so disruptive to require a indef block, without attempting to achieve consensus on the matter. IMO, the Arbitration Committee was formed to settle exactly this type of dispute instead of administrators over riding each other with their extra tool. In the future, if there is disagreement about a block, please bring it to the Arbitration Committee to be sorted out instead of using your tools in a manner that is controversial. FloNight♥♥♥ 11:12, 23 October 2007 (UTC)[reply]
  • The point I intended to make (and evidently is not clear) is that using your admin tools in a contentious manner needs to be avoided...whether it is blocking or unblocking. If there is substantial disagreement among administrators about whether an user should get an indefinite block or not, as there is in this case, it needs to be brought to the Committee where evidence is presented by all interested parties and a final decision that is going to stick can be reached. Using your administrative tools to block or unblock to make your views known is not the best way to handle the situation. It is confusing to users to see different admin interpret policy differently and use their tools in conflicting ways to supposedly enforce policy. After a case is accepted, we can take motions to lift a block if an user is blocked at the start of a case. If contacted by directly by an editor often we do it without a motion. FloNight♥♥♥ 19:29, 24 October 2007 (UTC)[reply]
    • Re-affirm my accept. Based on latest comments by parties, [2] [3] [4], they want ArbCom to review the situation and make a finding. FloNight♥♥♥ 15:06, 28 October 2007 (UTC)[reply]
• Accept. Charles Matthews 11:45, 26 October 2007 (UTC)[reply]
• Accept. James F. (talk) 07:13, 27 October 2007 (UTC) Reject, per Brad. James F. (talk) 09:31, 28 October 2007 (UTC)[reply]
• Accept, since the solution mentioned by Brad above does not appear to be permanent. Matthew Brown (Morven) (T:C) 20:52, 28 October 2007 (UTC)[reply]