Which option should this page include?

Option 1: "If the subject has denied such allegations, their denial(s) should be reported too."

Option 2: "If the subject has denied such allegations, their denial(s) should be reported too, while also adhering to appropriate due weight of all sources covering the subject."

Option 3: "If the subject has denied such allegations, their denial(s) should be reported too, while also adhering to appropriate due weight of all sources covering the subject and avoiding false balance."

Option 4: Remove all forms of this sentence. option added by Loki (talk) 02:41, 9 November 2022 (UTC)

Homeostasis07 (talk/contributions) 01:06, 9 November 2022 (UTC) question reworded for neutrality by DFlhb (talk) 02:03, 9 November 2022 (UTC); option 2 added by Homeostasis07 on 02:28, 9 November 2022

Context

Here's how this passage of WP:BLP evolved:

• Wording equivalent to option 1 was added on June 27, 2013

• It was copyedited to match option 1 on December 9, 2019

• It was changed to match option 3 on March 13, 2021

With a few bold edits and reverts in between. This RFC was preceded by the discussion above. A previous 2014 RFC addressed denials.

Note

This RfC concerns specifically the text found in WP:BLPPUBLIC. General BLP policies and related issues are set forth througout the BLP page. SPECIFICO talk 16:45, 9 November 2022 (UTC)

Survey (RfC on Denials)

• Option 1: Option 3 was added to this policy page on February 11, 2021 by a user who also linked to the essay WP:MANDY. This included MANDY's misinterpretation of the WP:FALSEBALANCE policy, which relates solely to "Conspiracy theories, pseudoscience, speculative history, or plausible but currently unaccepted theories" and how they "should not be legitimized through comparison to accepted academic scholarship." This in no way corroborates accusations of denials being excluded, or that the mere mention of a denial equates "false balance". Homeostasis07 (talk/contributions) 01:06, 9 November 2022 (UTC)

  Please note that this !vote incorporates a misinterpretation of the scope of WP:FALSEBALANCE. That passage weighs against the idea that every minority view or extraordinary claim needs to be presented along with commonly accepted mainstream scholarship as if they were of equal validity, and "mainstream scholarship" covers many of the "allegations or incidents" that BLPs would prefer to dispute and issue rebuttals for. The list that begins with "conspiracy theories" and "pseudoscience" is a set of examples - and there are also a number of BLP rebuttal/denials that rely on conspiracy theories or other WP:FRINGE claims. Newimpartial (talk) 16:40, 9 November 2022 (UTC)

• Strong support 1, oppose 2, strong oppose 3 & 4. We should explicitly favor including denials for allegations of crime or misconduct. That doesn't mean we give the same word count to denials and accusations; the denial needn't be more than a few words. But it's crucial to include denials for these allegations. Omitting denials will unduly bias readers towards believing the accusations are true, as scholarly sources show (see this whole reply: [1]). This is about respecting basic human dignity, and shouldn't be softened. DFlhb (talk) 01:17, 9 November 2022 (UTC)

The very lead of WP:BLP is also explicit that WP:V, WP:NPOV, and WP:NOR strictly apply (emphasis not mind). Saying so again here is unnecessary, excessive, and may be confusing. I also endorse Homeostasis's view of FALSEBALANCE, which is completely misapplied. Options 2, 3 and 4 will likely lead to denials being inappropriately excluded; denials for serious allegations are almost always due, but, as User:Masem hints at, that won't be obvious to all. DFlhb (talk) 02:29, 9 November 2022 (UTC)

• Option 3. WP:NPOV is a core content policy. We should not subvert it even in such cases. It is important to remind editors of that and to clarify that, when handled properly, there is no contradiction between that and this: including denials does not mean focusing on them to the point of a false balance. —David Eppstein (talk) 01:38, 9 November 2022 (UTC)

  Additional comment. My opinion above was formulated at a time when the RFC effectively only asked whether we should change from option 3 to option 1. Someone appears to have changed the bold text on my opinion to make it correspond to the newer 4-option version of the RFC. It is accurate that I prefer option 3 but I also think that making so many options confuses the issue, again, after it has already been more than adequately muddled in earlier discussions. —David Eppstein (talk) 04:50, 9 November 2022 (UTC)

  Option 4 also acceptable per XOR'easter's reasoning below. —David Eppstein (talk) 16:06, 12 November 2022 (UTC)

• Non-neutral RFC. This is not an "amendment". It has been stable on this page for two years now. The current wording is what you're framing as the "amendment", and what you're framing as the status quo is the actual amendment you want to make. I oppose this amendment and support the status quo but I think this RFC should be entirely remade, because it's currently both non-neutral and confusing. Loki (talk) 01:44, 9 November 2022 (UTC) Struck due to RFC rewording, actual !vote below. Loki (talk) 02:13, 9 November 2022 (UTC)

The text in the opening statement was as neutral as I could make it, given the circumstances on this talk page the past month. Opening statement has been amended further still to address any concerns. The original prose "If the subject has denied such allegations, their denial(s) should be reported too." had been incorporated into this policy for almost a decade, as evidenced by DFlhb, via both an RfC and numerous talk page discussions. This is the community-approved status quo version. Per both WP:ONUS (policy) and the WP:SILENTCONSENSUS essay you cited when reverting the policy, the amendment to the sentence has no mandate, since it has been disputed by at least 6 different users over the past month. Homeostasis07 (talk/contributions) 02:14, 9 November 2022 (UTC)

• Option 3 > 4 > 2; oppose 1: As David Eppstein says above, WP:NPOV (and WP:V) is a core content policy, so the actual policy is per Option 3 regardless of what this page says. Option 3 is therefore a clearer expression of this policy than Option 1 and having to correlate it against WP:NPOV and WP:V separately. Loki (talk) 02:13, 9 November 2022 (UTC) amended with ordering Loki (talk) 02:41, 9 November 2022 (UTC)
• Alice in Wonderland RfC - this RfC presents language that has not been stable on this policy page since the beginning of 2021 as though it were the current language, while depicting the language that has been stable on this policy page since February 2021 as though it were now an amendment. It seems designed to convince readers that down is up and up is down, and to vote accordingly. Newimpartial (talk) 01:53, 9 November 2022 (UTC) new !vote below Newimpartial (talk) 14:28, 9 November 2022 (UTC)
• Option 1. Between the first three versions, I would pick the shortest one mainly for the sake of conciseness. I don’t see the additional language adding much, because policies like WP:Undue will apply to some extent regardless. The added language really doesn’t seem to make much difference, it’s certainly not a license to delete any denial completely, given that it says to include denials “while" (not "if there is a way of") giving due weight and balance. I don’t have any opinion yet about which of these versions is the status quo, I only have an opinion about which version is better. Anythingyouwant (talk) 02:12, 9 November 2022 (UTC)
• It depends Personally I think the additional text is correct, but in alignment with a reading of UNDUE that recognizes that that a short denial is nowhere close to creating a false balance or attempting ot give equal weight. That seems to be essential to all this. --Masem (t) 02:15, 9 November 2022 (UTC)
• Option 1, then 2 then 3. If a subject has denied an accusation, that fact should always be included. Beyond that our policies and guidelines like undue and no false balance of course apply. I don't see a need to reiterate them here, much less leave open a possible interpretation that they can be used to exclude a simple denial in some cases.--agr (talk) 03:21, 9 November 2022 (UTC)
• Option 3. If an allegation is DUE it should have garnered sufficient discussion in RS for it to be presented NPOV. If the subject denies the allegation, one would expect that to receive satisfactory coverage as well, but if it doesn't then we don't need to report it. And MANDY doesn't link to WP:FALSEBALANCE, it links to the article false balance, which is pretty much consistent with the NPOV section WP:BALANCE. JoelleJay (talk) 03:46, 9 November 2022 (UTC)
• Option 3. Per NPOV as explained by David Eppstein, Loki, and JoelleJay. --Hipal (talk) 03:53, 9 November 2022 (UTC)
• Comment I don't think I like 2 or 3 because I think some editors will read those to mean, exclude if weight or false balance would be an issue. The concern about false balance and weight are legitimate but we should be clear that exclusion is not the way to address those potential problems. I'm going to hold off on !voting because I think 1 is the only option that makes it clear the denial should be included while the others, as written suggest only if inclusion wouldn't create false balance/weight issues. With some word smithing to make it clear that weight etc should be addresses by how not if the denial is included then I would say any of the 3 are acceptable. I'm holding off !voting in the event such changes are suggested. Springee (talk) 04:06, 9 November 2022 (UTC)
• Option 3/Option 2 - Either of these is correct (slight preference for 3 which is clearest), giving clear guidance that the denial should be appropriately covered without giving it undue weight. Option 1 is acceptable but lacks sufficient guidance on treating per WP:NPOV. Oppose option 4 which certainly allows editors to ignore WP:NPOV. Sirfurboy🏄 (talk) 09:31, 9 November 2022 (UTC)
• Option 3 Since there are some special rules that apply to BLPs, it's useful to remind editors that WP:DUE and WP:FALSEBALANCE are still in full force on BLPs. NightHeron (talk) 10:04, 9 November 2022 (UTC)
• Option 1 - in such cases, where RS or BLPSELFPUB sources, state the denial - we must do so. There is no case we can be absolutely sure he didn't do it; and if he explicitly denies it, we should say so - and even go slightly on the subject's side of UNDUE. #2 and #3 are likely to be abused when the subject denies the accusations, but reliable sources all think he did it. Animal lover |666| 10:47, 9 November 2022 (UTC)
• Option 1. If allegations are included in an article that a living person has committed a crime, then we must always include the subject's response to the allegations whenever that response is known. If we can only source the response to unreliable or partially reliable sources we must include with attribution (e.g. The Daily Mail reported ...), although if the only coverage is in such publications then always reconsider whether inclusion of the allegations is DUE. Options 2 and 3 imply that it is sometimes permissible to exclude the subject's response, but it is not. Thryduulf (talk) 11:17, 9 November 2022 (UTC)

  Thryduulf, this RfC is not particularly about criminal allegations. Part of your vote also makes it seem like you support considering WP:DUE when deciding whether to include the response.

  Since we have a different subsection about criminal allegations (WP:BLPCRIME), I'd support a brief statement about including pleas/denials with roughly the same language that is endorsed here. People looking for our policies on describing criminal allegations are unlikely to look for an important caveat in §Public figures. Firefangledfeathers (talk / contribs) 13:10, 9 November 2022 (UTC)

• The section being discussed (WP:PUBLICFIGURE ) is about potentially-defamatory statements; though this includes crime. And I fully agree with your second paragraph. DFlhb (talk) 14:46, 9 November 2022 (UTC)

• (edit conflict) In which case this RFC is badly presented as it mentions only "denials" and follows a section explicitly talking about crimes, and there is nothing in the proposed text including or excluding any type of allegation. Regardless, it whenever allegations (of any sort) are included in article it should always be a requirement to include the subjects response to them. DUE comes into play in deciding whether to include the allegations, in how much space to give the response to the allegations, how they are presented (e.g. quotes vs just a statement) and whether or not to include responses to the response, but the response itself should always be present in some form. Thryduulf (talk) 13:49, 9 November 2022 (UTC)

  Why? If we have a hundred strong sources that Richard B Spencer is a neo-Nazi, why do we have to include his denial? This is different from what we would do in any other situation, including if a separate person denied the allegations.

  I agree that a subject's denial of allegations definitely ought to be given more WP:WEIGHT than normal, but not infinite weight. Loki (talk) 17:31, 9 November 2022 (UTC)

  If we have a 100 strong sources saying a person is a neo-nazi then obviously the allegations are DUE for inclusion and deserve more weight and coverage than the subject's denial of those claims but their denial should still always be included. It doesn't need to be more than a single sentence but it does need to be there. If you describe that as infinite weight (I don't) then yes, it does deserve that. Thryduulf (talk) 17:37, 9 November 2022 (UTC)

  But why? What's the reasoning here? You keep on asserting that we must always include denials, but since this is an RFC, the weight of your vote is not the mere fact of it but the reasoning behind it. Loki (talk) 17:41, 9 November 2022 (UTC)

• Option 1 As long as it is fully attributed If someone denies an accusation we shous include that. Slatersteven (talk) 13:36, 9 November 2022 (UTC)
• Option 1 - the accused's response to allegations is always WP:DUE. Also, it's a basic tenet of journalistic ethics to report the accused's response to accusations when accusations are reported, so every RS will always report the denial, so it'll always be DUE. There is no way around it. The reason is because of obvious basic principles of fairness. If someone is accused of something, of course it matters whether the person admits or denies the allegations. Option 2 is OK as a second choice; I prefer the shorter version. Opposed to Option 3 because WP:FALSEBALANCE isn't applicable to this issue per others' comments above. (Also I agree we should not link the Mandy essay because I don't think it explains this issue well.) Levivich (talk) 14:05, 9 November 2022 (UTC)

  So if RS do not report a BLP subject's response to something (which may or may not otherwise be considered an "allegation"), should a denial be included in an enwiki article based on ABOUTSELF sources? Please see the second RfC, below. Newimpartial (talk) 14:09, 9 November 2022 (UTC)

  Does that ever happen in real life? RSes report an accusation, the accused denies it in public, and the RSes don't report the denial, so we only have the primary ABOUTSELF source available? I kinda don't believe it happens. I think if sources don't report the denial, then they're probably not really RSes, or it's only because we're in the short time period between when the RSes publish the accusation and when the RSes publish the denial. Is this a real issue? Levivich (talk) 14:16, 9 November 2022 (UTC)

  This often happens when it is the media itself making the accusation of part of the larger group that makes it. The dangers with modern journalism which have taken a more proactive role in their reporting Masem (t) 14:22, 9 November 2022 (UTC)

  Examples? Levivich (talk) 14:38, 9 November 2022 (UTC)

  Additionally, if the allegations are too recent for the person accused to have had chance to respond then it is going to be exceedingly rare for the allegations to be DUE for inclusion. In those rare cases where inclusion that soon is DUE, we should accompany them with a statement along the lines of "((as of|date)) $subject has not responded to these allegations". Indeed, we should include something like that whenever the subject has not responded, regardless of how long they have had to do so (which includes reporting explicit "no comment" responses). Thryduulf (talk) 14:25, 9 November 2022 (UTC)

  The problem I have with this is in determining what counts as an allegation. Some of the anti-MANDY activists would undoubtedly seize on this principle to insist that any reported incident that might be seen as reflecting negatively on a BLP subject be accompanied by a subject has not responded blank space to "insert denial here", which would be an absolute catastrophe for our articles on people who are uncontroversially documented as doing controversial things. Newimpartial (talk) 14:34, 9 November 2022 (UTC)

  which would be an absolute catastrophe for our articles on people who are uncontroversially documented as doing controversial things. how and why would this be a "catastrophe"? What is the justification for not reporting this? Thryduulf (talk) 14:49, 9 November 2022 (UTC)

  The justification for not reporting this is that the high-quality reliable sources on which our articles are based so not typically do so (where they do, we probably should). Wikipedia is supposed to be an encyclopaedia, not a small-town local newspaper. Newimpartial (talk) 14:57, 9 November 2022 (UTC)

  Imagine a celebrity shot someone. For the sake of argument, it's on camera and in lots of newspapers, clearly WP:DUE. Regardless of whether we have to say "[CELEBRITY] denies the allegations" if the celebrity has in fact denied the allegations no matter how well-verified they are, it's obviously absurd to say that if the celebrity has not in fact denied the allegations.

  The more important problem here is that on a page like David Icke, it's problematic even to say that he doesn't think he thinks he's a conspiracy theorist. (Of course he doesn't, we report that he believes what he says and that oughta make it clear that he doesn't think he's a conspiracy theorist.) If we go beyond that to say, for instance, that Icke might say nobody's proven the queen wasn't a lizard-person, then we're just pushing WP:FRINGE beliefs in the guise of a denial nobody has actually made. Loki (talk) 17:56, 9 November 2022 (UTC)

  Levivich, you may not have seen the many, many times I have seen anti-MANDY activists insist that an RS characterization or a well-documented event counts as an "allegation" or an " incident", and therefore that a SELFPUB rebuttal must be included in the BLP article (and also typically that the RS characterization or event be taken out of wikivoice and given the same standing as the denial). This isn't a hypothetical case. Newimpartial (talk) 14:39, 9 November 2022 (UTC)

  Of course I haven't seen that, which is why I'm asking for concrete examples. I also don't know what an anti-Mandy activist is. What I do remember is seeing people cite Mandy as justification for removing Julian Assange's denial of rape accusation from the article. That was wrong, it should have been included, and in the end, it was, in part because of this line in BLP policy. So option 1 unless someone can show me a slew of real examples where option 1 posed a problem. Levivich (talk) 14:47, 9 November 2022 (UTC)

  Here is a long but otherwise typical example from Talk:Lauren Southern; she or her people visit the page regularly to insist on the exclusion of well-sourced characterizations or the inclusion of SELFPUB denials. The instance in question was settled in relation to the existing (Option 3) language by including a denial that was picked up by RS and excluding other denials that were SELFPUB; had Option 1 been in place, only WP:IAR could have been deployed to prevent the inclusion of FRINGE, self-serving denials. Newimpartial (talk) 15:03, 9 November 2022 (UTC)

  Denails are not the same type of arguments as FRINGE. The latter deals with pseudoscience and conspiracy theories that have been soundly rebutted by academics and others, meaning WP should never even consider fringe topics as being legitimate concepts. Treating denials of yet proven allegations as FRINGE is immediately hostile to BLP, and cannot be an avenue we take. Masem (t) 15:14, 9 November 2022 (UTC)

  What do you mean by yet proven allegations? I have seen you say that enwiki should wait for more than ten years of scholarship before accepting characterizations of events and of public figures in wikivoice - if implemented, that would result in a lot of WP:FALSEBALANCE about events (like the 2020 US election and the January 6 Capitol Hill debacle) to which WP:FRINGE most certainly applies. A "false flag" rebuttal of the event published by a Capitol Hill protester on Facebook is not DUE for inclusion in a Wikipedia article just because it exists, even though it may be SELFPUB by a BLP. Newimpartial (talk) 15:24, 9 November 2022 (UTC)

  We are talking like when BLP are called racist or similar language, which are things that cannot be proven by any authoritative source. We can only wait to see long long-term coverage handles such claims, if they stick over the test if time. This isn't covering something like the Jan 6 false flags; it has been determine by courts, reputable academics and politicians as a hostile attack on the Capital, so any denials that claim peaceful protests do fall under FRINGE so we'd not include the false flag complaint. That's also not an allegation directed at the BLP. Masem (t) 15:51, 9 November 2022 (UTC)

  It is for BLPs who are known (reliably sourced) to have been part of the attack. Also, I don't think other editors agree with you that we can't identify a public figure as "racist" or "alt-right" without 10 years of scholarship, and I can't understand the categorical distinction you are trying to make between allegations and FRINGE claims - of course most denials of accusations don't involve FRINGE claims, but a number of them do. Newimpartial (talk) 16:00, 9 November 2022 (UTC)

  If a Han 6 protestor tried to deny by claiming the false flag "it was a peaceful protest" route, and importantly not challenging their involvement. that would be following the established FRINGE aspect related to Jan 6, and thus properly wed minimize that denial. On the other hand, if that person made a denial that was more personal to them like "I wasn't there, you have me mixed up" or "I was way off from the building cheer others on" that would be nonFRINGE denials that should be include. There is a subtle but significant distinction here. Masem (t) 16:08, 9 November 2022 (UTC)

  The Southern case is an interesting one. Should Wikipedia cover self declared change in view/perspective/heart etc of it's not in a RS. That's a hard one since I can see why it would be similar to a denial but it's not the same thing since this isn't a case odd disputed claims. The same might be true of an MD who changes their perspective on a controversial medical topic. I don't think denial applies in such a case. It might be good to have something that does but we currently don't. It may also be hard to decide when such a disavowment is genuine ( in which case I think we should cover it) vs when it's a wolf in sheep's clothing. That's a case where I would probably say no if we only have a SP claim but yes if the claim is published even in a normally poor source. Springee (talk) 15:32, 9 November 2022 (UTC)

  Graham Linehan is another figure who has, along with his followers, tried to edit his page to remove what he regards as false "allegations" and to include poorly-sourced denials. Option 1 would make it much easier for them, and their allies in editing that page (some of whom are among the principal authors of the anti-MANDY essay), to do so. Newimpartial (talk) 15:12, 9 November 2022 (UTC)

  We're not talking about the same thing. "Alt-right", "white supremacist", and "anti-transgender activist" are not allegations or accusations, they're characterizations. They're labels. The text at issue in this RfC is from WP:BLPPUBLIC and talks about allegations and accusations. The example given in BLPPUBLIC right before the line being discussed is a politician alleged to have an affair. The difference is that an allegation or accusation is about something someone did, not something they are. If the politician denies having the affair, we should include the denial (as will all RSes). If a BLP subject disagrees with the labels an RS applies (like "alt-right" or "far left" or whatever), that's a whole different issue (MOS:LABEL). Levivich (talk) 15:28, 9 November 2022 (UTC)

  I have made precisely the argument you are making here, about the scope of allegation or incident, but the anti-MANDY faction typically won't hear of it. Another example, where I don't have a diff at hand, was at one of the Florida politics articles where certain editors were arguing that a self-published rebuttal of something that was simply RS reporting, not an allegation or incident needed to be included because of the "right of reply". This kind of argument is being made constantly to relativize facts that are part of consensus reality and characterizations that are universal within high-quality sources. Move to Option 1, and I can guarantee that these arguments will proliferate and become more successful, even though a few well-intentioned Option 1 !voters are only thinking about actual unproven allegations and incidents. Newimpartial (talk) 15:56, 9 November 2022 (UTC)

  Allegation of kabobs nearly always result from what someone did, and more often the cumulation of things they did, and are meant to harm the person's reputation, so we should see them similar to allegations with legal ramifications. Masem (t) 16:01, 9 November 2022 (UTC)

  I really hope the use of "kabobs" here was intentional. Newimpartial (talk) 16:05, 9 November 2022 (UTC)

  The dish was alleged to be a kabob,^[1][2][3] although the chef denied it.^[4][5][6] Levivich (talk) 16:10, 9 November 2022 (UTC)

  The LABEL said it was a kabob. Newimpartial (talk) 16:51, 9 November 2022 (UTC)

  If an allegation is unproven then we cannot report it as fact in Wikivoice, regardless of what anybody else says or doesn't say about it (per WP:NPOV). Thryduulf (talk) 14:52, 9 November 2022 (UTC)

  I am not taking about unproven allegations, but rather "proven" ones (by whichever standards of evidence are relevant to the claim). Newimpartial (talk) 14:59, 9 November 2022 (UTC)

  User:Newimpartial, you are talking about contentious labels (not allegations), which are covered by WP:CONTENTIOUS and WP:DUE. Contentious labels are additionally covered by WP:FRINGE if they have anything whatsoever to do with scholarly or scientific topics. This is very clear, and is existing policy.

  The denials phrase explicitly says "such allegations" (referencing the "messy divorce", "affair" examples), and emphatically doesn't apply to contentious labels.

  The reason for your opposition to option 1 seems to have been founded on a misunderstanding: you believe your preferred wording would enable us call Lauren Southern "alt-right" in Wikivoice, for example, and that option 1 would prevent us from doing so.

  Wrong! Lauren Southern was already called "alt-right" in Wikivoice back when this text followed option 1!. So was Mike Cernovich, also before option 1 was changed.

  You're proposing a potentially-dangerous watering down of something that is solely meant to protects innocent or vulnerable people (by letting them defend themselevs from potential defamation). This won't have any effects on Lauren Southern, trust me. But it will harm the dignity of many, including racial, ethnic, gender minorities, who are more likely to be covered sensationalistically by the media, and for whom allegations of crimes are more likely to be believed. Obviously, this applies to everyone else, too.

  (I hope you'll read this comment carefully, as I do yours, to minimize the need for back-and-forth clarifications, and that you'll address the substance of it; I take responsibility for not "getting" the core of our disagreement earlier, which could have saved both of us a lot of time.) DFlhb (talk) 16:09, 9 November 2022 (UTC)

  Please note that I linked to a discussion on Talk:Lauren Southern that was settled while the Option 3 language was in place and following the logic of that language. That isn't a straw man argument, and the fact that pages have been edited to acceptable outcomes without the support of optimal policy language isn't really an endorsement of suboptimal policy language IMO. Also, please don't use gotcha (Wrong) markup when you strawman (seems to have been founded in a misunderstanding) my argument - by doing so you will not minimize the need for back-and-forth clarifications in spite of any intention to the contrary.

  The fact remains that certain editors consistently have, and will continue to insist on a "right to reply" for characterizations applied to BLP subjects and that they have, and will continue to deploy this passage of policy text to do so. The fact is that this policy text is not limited to accusations of legal or clear ethical wrongdoing, nor is it limited to accusations that are "not yet proven" - in its plain meaning, it continues to apply even after a court case is settled or after RS scholarship has been written, and enhancing the "right of rebuttal" for UNDUE content from BLPs is the opposite of an encyclopaedic objective. Newimpartial (talk) 16:33, 9 November 2022 (UTC)

  Apologies in advance for the length. I've read both that discussion and the page history. It is common on these talk pages that such contentious terms are disputed, especially by various IP or red link users, and occasionally by subjects; Wikipedia explicitly encourages that, and no policy change can prevent disputes. Certain editors insisting on certain changes, or disagreeing with you on policy interpretations, is inevitable, no matter what any policy says.

  I somewhat agree with your argument which begins with The fact is that this policy text is not limited [...]: but no option in this RFC addresses it substantively.

  Denials may still be warranted:

  • After court cases are settled (see the Innocence Project, a primary source which is clearly undue if it's the only one raising doubts and no WP:RS has covered their claims; yet still merits inclusion)
  • After scholarship has ben written (this applies more to politicians and major figures, right? The way they deny allegations is likely noteworthy in-and-of-itself; their denials could be grotesque, or discredited by later scholarship (even if that scholarship doesn't mention the denial), which I'd consider noteworthy, to give readers an indication of a subject's trustworthiness, reliability, or character). And to bring up a separate but related point: MEDRS may say some orgs are, say, transphobic, but I doubt they'll call individuals transphobic, no? So I'm unsure how that would apply to BLPs; I'd appreciate examples.
  • There is no reason to blanket-ban self-published denials, which may be due; let's say multiple WP:RS mention that a denial exists, but not its actual claims; it may well be appropriate to include the claims.

  My point is that common sense prevails here; there's nuances to everything, and we should avoid excessive detail here when specific issues are best settled on article talk pages. Trying to enshrine a "one-size-fits-all" rule for when denials should or shouldn't be included isn't within the purview of a Policy, and wouldn't prevent any article disputes (the more detailed a policy is, the more material there is for Wikilawyers to twist). We should Emphasize the spirit of the rule (as WP:PG states), and trust that editors may make mistakes, but will eventually get it right. WP:PG states: Verbosity is not a reliable defense against misinterpretation. We should use direct, concise writing. We should use footnotes [...] for further clarification, not inline policy changes.

  The wording I support says "should be reported", not "must". That's clear enough to me to allow for common-sense to prevail, which it should, as opposed to a top-down "edict". I've argued why I think option 3 is too ambigious and will likely lead to misinterpretations, which could promote talk page disputes rather than prevent them. Right-to-rebuttal should be determined on a case-by-case basis on article talk pages, as is currently done. It works well. DFlhb (talk) 17:33, 9 November 2022 (UTC)

  My point remains that your "should be included" language, in the absence of the caveats that have been present in the policy text since 2021 will (almost by definition) encourage the interpolation of self-published and marginally reliable "rebuttals" that are UNDUE and self-serving. I also disagree that it is the place of Wikipedia to platform self-published denials after conviction until independent RS have covered them, and I don't see the point in including denials self-published by public figures in cases where these are ignored by both contemporaneous RS and subsequent scholarship. (On your transphobia example, This online transphobia has been spearheaded by former comedy writer Graham Linehan [2] is undoubtedly a BLP claim whether or not it is an "allegation or incident", and regardless of whether it is a statement to which WP:LABEL would apply).

  For the record, I do agree that once RS have lent WEIGHT to a denial or rebuttal it may be appropriate to source additional detail from the ABOUTSELF source, but that isn't a denial sourced only to ABOUTSELF material in the sense of the other RfC . Newimpartial (talk) 17:55, 9 November 2022 (UTC)

  Here's another way to look at the issue: if we carve out a domain in which SELFPUB statements must be included regardless of DUE and NPOV considerations, then the question will once and forever be, "how large is this domain?" In this scenario, editors will have a strong incentive to WP:GAME the domain to win content disputes by expanding the scope of what count as "allegations or incidents". Keeping the current language doesn't entirely prevent the problem from arising, but at least it doesn't make it worse (which will be the result of Option 1). Newimpartial (talk) 17:27, 9 November 2022 (UTC)

• Option 3 first, Option 2 second; Option 4 before option 1 - I fail to understand the objections to the WP:FALSEBALANCE hyperlink; there are certainly many denials or rebuttals offered by BLP subjects where the quality of sourcing that something actually happened as described is very high (including RS "scholarship") even though a BLP subject denies it.
• Many "Option 1" !voters appear to believe that the policy passage in question applies only to unproven allegations or incidents, as though once legally proven or documented in high-quality scholarship the supposed "right to rebuttal" would conveniently no longer apply. However, I see no justification in either the policy language itself or in the behaviour of anti-Mandy advocates on BLP pages to expect these editors not to use an Option 1 RfC outcome to shoehorn BLPSELFPUB and other poorly sourced FRINGE denials into BLP pages, and to insist that allegations and incidents that have been "proven" in court or in high-quality scholarship be presented out of wikivoice as attributed statements in FALSEBALANCE with the subject's denial. Newimpartial (talk) 14:27, 9 November 2022 (UTC)
  • Allegations/accusations that are proven are no longer allegations or accusations. But to take an obvious example: if someone is prosecuted for a crime and pleads not guilty and is convicted and nevertheless maintains their innocence, we would and should report that they nevertheless maintain their innocence, and furthermore, every RS would also report this. Even if an entire nation were accused of something (like war crimes) and denies it, we should report the denial. What is the other option, anyway? That we omit the denial? Why would we do that? The truth is, these three options are really the same. A denial will always be reported by RS and so it will always be DUE for inclusion. Also, what the subject of a allegation/accusation/criticism/complaint/controversy has to say about it will always be relevant, obviously, just as it will always be relevant who is making the allegation/etc., just as it will always be relevant what the applicable authorities (eg courts) decided about it. We always "hear from the accused." This is just how the world works, and I really am straining to understand what actual real-life examples we are concerned about here. The right to rebuttal is basic journalism ethics, and it applies even to convicted murderers on death row. Levivich (talk) 14:42, 9 November 2022 (UTC)

    Please see my comments in the discussion of your !vote. I have seen plenty of cases where BLP subjects deny elements of consensus reality and where these denials are not picked up at all by responsible media; there are also plenty of cases where anti-MANDY activists pick up marginally-sourced (but not fully SELFPUB) denials as grounds to place consensus reality and FRINGE perspectives on equal footing. If you are imagining the scope of this language as only covering allegations of crimes and similar wrongdoing (affairs, etc.), you are experiencing a failure of imagination in relation to the way the DENIALS text is already used onwiki, before any watering down of the current language. Newimpartial (talk) 14:46, 9 November 2022 (UTC)

    If you are imagining the scope of this language as only covering allegations of crimes and similar wrongdoing this RFC wouldn't clarify the scope in any way, it just increases the bar for including all denials, including criminal ones. I'd strongly support an RFC that actually addresses what you point out; this isn't it. DFlhb (talk) 17:47, 9 November 2022 (UTC)

  The corralary that many option 2 or 3 supporters may feel inclusion is a given but they are concerned regarding weight given to a self published denial vs RS coverage. Others may think those options mean the denial itself can be viewed as Undue. At least I think this it's getting to the core concern. Springee (talk) 18:37, 9 November 2022 (UTC)

• Option 1– Options 2 and 3 are unclear, suggesting that denials can be excluded if someone argues that including a denial would be undue or ‘false balance’. Denials should always be included; the total amount of coverage of denials/accusations may be subject to considerations of due and balance.Sweet6970 (talk) 15:28, 9 November 2022 (UTC)
• Option 1 - while I agree with what option 2 says, I don’t think the policy needs to explicitly say it. DUE/UNDUE always applies, so stating that it also applies to denials is simply unnecessary instruction creep. Blueboar (talk) 16:59, 9 November 2022 (UTC)

  The problem is, if we don't include it here, we get a lot of denial-maximalists who insist on including all denials for any allegation no matter how well proven it is. Loki (talk) 17:34, 9 November 2022 (UTC)

  Presumably by allegations you mean "transphobic", not "cocaine user"? I've asked someone else before, but I'd appreciate an explanation for why WP:CONTENTIOUS and WP:DUE don't already cover the former. The BLP section we're discussing is specifically about things like the latter, as clarified by the examples it gives. I don't see how watering down the wording would help you. No policy change can prevent editors misinterpreting policy. DFlhb (talk) 17:43, 9 November 2022 (UTC)

  (edit conflict) If a person verifiably denies an allegation against them, we should report they deny it. How well proven it is is irrelevant. Per NPOV we present the relevant facts in proportion to their DUE weight, without making value judgements, and the let readers make up their own minds. A person's denial of an allegation against them is always DUE. What differs is how much weight we give it with then article relative to the claims, but it is never zero. Thryduulf (talk) 17:44, 9 November 2022 (UTC)

  But that's not in here, or WP:DUE. Nothing is always WP:DUE. WP:DUE is a matter of sourcing, not content. Loki (talk) 17:58, 9 November 2022 (UTC)

  I suspect that the closer may have difficulty parsing the difference between Option 1 !votes based on the premise that DUE applies anyway, so it need not be mentioned, and Option 1 votes assuming that all denials should always be included so DUE isn't relevant. Newimpartial (talk) 17:37, 9 November 2022 (UTC)

  DUE does matter - it is just that mentioning a denial is always DUE. Even if that wasn't the case, the practical difference between your options is zero, so it doesn't matter. Thryduulf (talk) 17:46, 9 November 2022 (UTC)

  Always due for allegations, not for contentious labels (which WP:PUBLICFIGURE doesn't cover). This RFC blurs the two, and I'm highly distressed that a consensus for 3 (or god forbid 4) would apply to both, not just the latter, despite many voters here seemingly only addressing the latter. DFlhb (talk) 17:56, 9 November 2022 (UTC)

  But you have editors like Masem arguing that these are not separate issues - it isn't only Option 3 !voters who see a connection. Newimpartial (talk) 18:03, 9 November 2022 (UTC)

  I believe that is a minority view, but regardless, I feel this RFC should actually address whether there is a connection (which IMHO would address your arguments, mine, and others' far better). It seems that's the main point of contention that led to the RFC in the first place; not the wording. DFlhb (talk) 18:11, 9 November 2022 (UTC)

  It may not matter to how the resulting text of this policy is written, but it does matter to what it ought to mean. If some Option 1 !voters believe rebuttals can be excluded based on DUE (e.g. in edge cases, because DUE always applies) and other Option 1 voters believe that all self-published rebuttals must always be included, then one of the two groups will be dissatisfied with the way the resulting language is interpreted (or possibly each of them may be disappointed part of the time). Newimpartial (talk) 18:01, 9 November 2022 (UTC)

Agreed. I think if none of us even agree on what the options would mean. DFlhb (talk) 18:08, 9 November 2022 (UTC)

• Option 3 Per WP:NPOV and WP:V, strong oppose of option 1 for all the reasons people have already mentioned. -- LCU ActivelyDisinterested ∆transmissions∆ °co-ords° 18:06, 9 November 2022 (UTC)
• Comment: should we blow it up? the RFC was (very reasonably, but unilaterally) started by one editor; none of the options or wordings were workshopped beforehand.

There seem to be very strong disagreements, not even on what option to pick, but on what these options mean, and what their scope is. Does it apply to crimes? to contentious labels like "racist"? (I don't think the section we're discussing even covers this; yet others seem to solely focus on it. It seems to be a major point of contention, yet the RFC doesn't even address it). It feels like we're all arguing about different things.

I'd like to borrow from User:Newimpartial's idea below, and propose we pause this RFC for now, and, together, workshop a new RfC that reflects everyone's consensus. I'm thinking of a multi-step RFC, again borrowing from Newimpartial's idea, where we have individual subsections about "how should the policy apply?" (does it apply to crimes? contentious labels? misconduct? how should self-published denials be treated? under what criteria is it undue? should such criteria be listed inline, or in an endnote?). The current RfC would be binding for at least a few months, but I don't think it would make anyone happy. It may indeed be wiser to start with seeking consensus on application and working out way back to wording. Highly interested in thoughts. DFlhb (talk) 18:08, 9 November 2022 (UTC)

I'd be fine to do that; I'd also be fine for my RfC below to be euthanized prior to the worshipping process for a new RfC/new RfCs. Newimpartial (talk) 18:11, 9 November 2022 (UTC)

I'll give the eulogy, you play the organ. DFlhb (talk) 18:17, 9 November 2022 (UTC)

I'm not a fan of killing off a RfC after pale have replied but I think out makes sense here. It seems we now at least agree on the question. Better phrasing should fall out once we have a consensus on that. It also means of there is ever a disagreement editors can point to the talk page discussion to understand intent. I would suggest pinging all participants both now and once a new Rec launches. Springee (talk) 18:42, 9 November 2022 (UTC)

I've started a new subsection to workshop a new RFC here, feel free to join. My comment above argues in favor of blowing up the current RFC, since it already seems headed to an impasse no matter how long it runs for, with editors disagreeing on what the options would mean. Note that there isn't yet consensus that a new RFC is warranted, it's still just a proposal; not enough editors have chimed in on it yet.

Pinging users who participated, as per Springee's suggestion, I agree that's warranted. Homeostasis07, Newimpartial, David Eppstein, Loki, Anythingyouwant, Masem, agr, JoelleJay, Hipal, Springee, Sirfurboy🏄, NightHeron, Animal lover, Thryduulf, Slatersteven, Levivich, Sweet6970, ActivelyDisinterested. I think that's everyone. DFlhb (talk) 18:58, 9 November 2022 (UTC)

Thanks for the ping, I am happy with a new RFC as it's clear this isn't going to come to a consensus any time soon. Thryduulf (talk) 22:28, 9 November 2022 (UTC)

I'm also okay with pausing the RFC and workshopping now we have a better idea of what areas are contentious. Loki (talk) 18:59, 9 November 2022 (UTC)

• Yes, please This is only going to insitutionalize misunderstanding and discord. Would need to be very sharply defined and provably needed in addition to core policies, which IMO already give enough guidance. SPECIFICO talk 19:04, 9 November 2022 (UTC)

  I'm highly sensitive to the issue of discussing in a vacuum without taking into account how our changes relate to other policies, since any changes would have a large impact. How sharply-defined do you think this initial workshopping proposal is? I'd highly appreciate any input on how we could make it appropriately-sharp. DFlhb (talk) 19:16, 9 November 2022 (UTC)

  Sorry, @DFlhb: but I believe this RfC can serve an important purpose in any future talk page discussion. The primary purpose here is to address the issue of "implied consensus" that certain users raised in discussions above. One way or another, that can't be argued again after this. And the completion of this RfC does not impact any future RfCs, in my opinion. Things can be reworded in future RfCs to accommodate the result obtained here. Either way, this is an issue that needs to be resolved before substantive progress can be made. Homeostasis07 (talk/contributions) 02:06, 17 November 2022 (UTC)

• Option 1, when serious accusations are made against a person and they deny it, we should mention it. It can be very brief, but this is typically what is done in reliable sources. Yet, many proponents of WP:MANDY - which is what inspired the addition of the clauses found in options 2 and 3 - suggest such should be excluded even when it's reported in many sources because doing so is allegedly UNDUE, which brings Wikipedia out of step with reliable sources when done. More on this at WP:NOTMANDY. Crossroads ^-talk- 22:57, 9 November 2022 (UTC)
• Option 1, as I don't see why we would want to put anything in the way of adding reliably sourced denials. Often the recipients of damaging accusations are also the subject of media bullying and bias, and do not necessarily have access to the full range outlets to present their side of any stories. The only thing that putting obstacles in the way of using any available reliable sources will do is reduce the likelihood that the article will fairly reflect both sides of the argument per WP:NPOV. -- DeFacto (talk). 09:40, 10 November 2022 (UTC)
• Question What if the only source for the den[ial of] such allegations is in a self-published source (e.g., Twitter)? Wouldn't that be unduly self-serving in many cases? And what if the allegation also makes a claim about a third party (e.g., "The person who accused me is a known liar who always lies")? In any situation like this, policy would forbid us from including any details of the denial. Accordingly, none of the given options sound quite right to me. XOR'easter (talk) 13:31, 10 November 2022 (UTC)

  I don't believe so. It arguably is self serving but the key is "unduly". I view unduly statements/comments as ones that putting someone to above where they started. So the reputation of Dr X is at 0. Dr X is accused of abusing patients. Now their reputation is somewhere in the negative since even if we decide the allegation is false, it's not like we think more of Dr X for being accused. A denial may make us think the accusations are less likely to be true but they will never raise Dr X's reputation to somewhere above where it started. Dr X talking about his great therapeutic breakthrough would be unduly self serving since it would be a self published claim that tries to raise his reputation above it's starting point. Springee (talk) 13:38, 10 November 2022 (UTC)

  It might help if we separate denials from rebuttals. A blunt statement mentioning that the subject denies the allegation is purely an ABOUTSELF statement, allowed under SPS. But once you start going into the details - the reasons why the subject denies the allegation or his rebuttal of the allegation, we potentially cross the line into talking about third parties… where we would need a non-SPS source. Blueboar (talk) 14:22, 10 November 2022 (UTC)

  In some cases the denial is all that is required, certainly, but sometimes the comments with them are significant even if self-published. For example if Miss Smith accused Mr Jones of raping her at a party, there is a difference between Mr Jones saying "I was alone with Miss Smith, but we didn't have sex", "Miss Smith and I had consensual relations", "I have never met Miss Smith", "I didn't rape Miss Smith, it was Mr Bloggs" and "I was in a different country with Ms Doe at the time". How we should cover that depends on the context (including the details of what is alleged), so the only one-size-fits-all rule we can and should have is that when allegations have been verifiably denied, we include the fact of the denial even if it's only an an SPS. Thryduulf (talk) 17:54, 10 November 2022 (UTC)

• Option 1 i.e. return to version before the bold edit that added the clause. The addition could mislead editors into thinking that denials can be suppressed because someone deprecated the publication. I am aware of the previous RfC (11 to 1 in favour of "mention and/or link to the response/rebuttal"). I am aware of the attempt to abort this RfC. I will not respond to heckling. Peter Gulutzan (talk) 13:33, 10 November 2022 (UTC)
• Option 3 It is clear from the longevity of the clause that it is generally agreed a denial should be reported if given. That is an explicit clause overriding WP:DUE. However it is also evident some people view a denial as including in Wikipedia the content of the denial, and that is not gererally agreed. I think we need the extra clauses to deliniate the extent of the override - it only includes saying it is denied and having a citation. Anything more has to follow policies as normal. NadVolum (talk) 14:34, 10 November 2022 (UTC)

But if it is spelled out in other policies, do we actually need to repeat and spell it out in THIS policy? I think part of the argument against option 3 is that it is instruction creep. Blueboar (talk) 15:51, 10 November 2022 (UTC)

• Option 1 In a worst case scenario, if a denial is clearly false (remembering that this is primarily targetting people who claim not to be anti-vax or racist, rather that people accused of crimes), this will be evident to any reader if they look at the sources provided. However, if we leave out a denial, we risk causing significant harm to the subject, especially if not including their denial of the accusation might falsely lead the reader to assume it is true as only one side is given. On thst basis, the default position should be to include all denials even if editors agree that they should not be given much weight, with weight being an issue for local consensus rather than policy. - Bilby (talk) 19:25, 10 November 2022 (UTC)
• Option 1 I don't believe this was broke. Morbidthoughts (talk) 04:23, 11 November 2022 (UTC)
• Option 1 Not sure DUE is needed to be mentioned. However, I seems this whole thing has arisen by editors citing this section to deal with, as Levivich notes, labels or characterisations, which is covered by MOS:LABEL. It is easier to resolve allegations that someone did something or said something, whereas resolving how we describe someone is hard. As I noted in the section above, we should not link to WP:FALSEBALANCE or False balance. However, I think the issue of what label we give someone is too important for MOS and should be covered here. But it likely needs a lot of thought, and should be dealt with separately from other allegations, for which the existing policy was fine. -- Colin°^Talk 11:19, 11 November 2022 (UTC)
• Option 4 I don't see why we are carving out a policy exception for one particular type of content. If the denial is self-published, then WP:BLPSPS applies, and the language here amounts to an insistence that no denial can ever be unduly self-serving. We need the freedom to debate that where appropriate, not an edict forbidding us to do so. If the denial is reliably reported elsewhere, then we don't need any of these sentences at all, because WP:NPOV and WP:DUE would give sufficient reason to include it. Overall, I find it really weird that any particular type of content could be made mandatory at the policy level, and I don't like editorial discretion being overridden outside of break-glass-in-case-of-emergency emergencies. XOR'easter (talk) 15:32, 12 November 2022 (UTC)

  WP:BLPSPS applies for self-published statements in general. There was a 2014 RFC, and the consensus was that denials always merit inclusion, even if self-published. This RFC cannot 'overturn' that, since it focuses on wording and doesn't address that question explicitly. I think removing the sentence would just mislead people as to this consensus.

  And secondly, removing the passage, and relying purely on other policies (like DUE) would also introduce ambiguity, since some might argue that even a denial covered by WP:RS isn't due (since it's usually only mentioned in passing by news orgs).

  There are two issues: what the consensus is on how to treat denials (policy), and how to express that consensus (wording, which is all this RFC focuses on). Since this RFC is purely about the latter, and doesn't address the former; and since the 2014 RFC explicitly addressed the former; this RFC is simply malformed, and would constitute an improper 'overruling' of the 2014 RFC. I'll note that several of the people who called for this RFC expressed an intent to exclude some denials. DFlhb (talk) 17:12, 12 November 2022 (UTC)

• Option 4 otherwise Option 3. Our editing works best when we keep our core policies at the forefront and do the messy work of collaborating on how to apply them to the infinite variety of situations, sources, content, and considerations that apply to specific article content. The more we try to specify subsidiary or derived mandates, the less likely we are to find appropriate solutions to the actual issues and problems that arise in creating article content. Our overriding BLP, NPOV, RS, and V policies need to be applied case by case and we should not elevate any of the proposed wordings to ironclad rules that have the force of law. They are not flexible or detailed enough to be operationally useful. They're rules that sound nice in the abstract but it's not obviou how to apply them to the real-world editing issues that arise. SPECIFICO talk 02:32, 13 November 2022 (UTC)
• Option 1. The short and sweet direction. Pyxis Solitary (yak). L not Q. 14:18, 13 November 2022 (UTC)
• Option 1. If allegations are serious enough to merit inclusion in an article, readers are surely entitled to know whether the subject has denied them or not. Frankly, a bizarre question to ask. 'Due weight' could never justify excluding such a denial, so doesn't really come into it. AndyTheGrump (talk) 06:15, 15 November 2022 (UTC)
• Option 1. We need to do our job right. As a matter of fairness to living people who are harmed by gossip and rumors published by those so-called "reliable" publications who pick quarrels and provoke trouble, we should give space to the people in our articles when they issue credible denials of rumors and gossip. Ghost of Kiev (talk) 15:14, 15 November 2022 (UTC)
• Option 4 > 3 > 2 WP:NPOV, and by extension WP:DUE, is a core content policy. It is non negotiable. It's a fact that many denials are considered by the balance of our sources to be non-credible, and are not given any airtime. To require every denial to be included as option 1 states, especially in situations where our sources do not include them, is to contravene both the spirit and the letter of WP:DUE. If a denial of an allegation is only self-sourced, then to quote from DUE If a viewpoint is held by an extremely small minority, it does not belong on Wikipedia, regardless of whether it is true, or you can prove it.
  Option 4, the version that removes the sentence on denials completely, best fits within how we define and apply policy, as it allows for inclusion of denials when they meet DUE/NPOV. Option 3 and 2 are a close second/third, as they make it explicitly clear that any denial must be tempered against our core content policies. Option 1 unfortunately, as I said a couple of sentences ago, violates policy and encourages inclusion of denials regardless of how they are reported by reliable sources. To the editors who are making arguments on scope creep, I would remind them that not all editors are intimiately familar with every policy. In cases where it can be unclear how to apply conflicting policies, such as in this policy, it is better to be explict as to how those policies interact with each other. Not only does this prevent unnecessary repetitive discussions across many article talk pages and noticeboards, it also makes our labyrinthine policies clearer to new and otherwise unfamiliar editors. Accordingly I would see the extra text in options 3 and 2 as WP:NOTCREEP as it is a helpful additional instruction that succinctly states community consensus and avoids harm when the alternative (option 1) is unclear. Sideswipe9th (talk) 19:14, 18 November 2022 (UTC)
• Option 4 > 3 > 2. As explained above, we decided whether to include something based on WP:NPOV and WP:V; it is completely inappropriate to require specific inclusions without reference to those things, which form the core to our project. We cannot simply ignore sourcing and weight out of a desire to WP:RIGHTGREATWRONGS. As a reminder (which many people seem to be overlooking), this entire section is about public figures where the allegations are well-documented in multiple high-quality sources; at that point reporting them on Wikipedia has no real risk of harming them, since they are well-known and this is part of the core principle behind WP:PUBLICFIGURE - but giving undue weight to denials could potentially harm our neutrality and, by extension, our overall mission. Option 1 is against core policy (because it implicitly instructs editors to ignore NPOV and V and to include things with no regard for sourcing) and is therefore not something that can be implemented by consensus; even if the text were to say that, it would be inaccurate and denials could (and would) continue to be removed if they completely failed the requirements of WP:NPOV and WP:V. Some editors above have argued that "due weight" could never justify excluding a denial; if they genuinely believe that to be the case then the other options should also be acceptable to them, but the fact that we are here and arguing the point makes it clear that that is not the case. --Aquillion (talk) 20:03, 18 November 2022 (UTC)
• Strongly Oppose Option 3. The wording makes it seem highly subjective. Styx (talk) 05:49, 20 November 2022 (UTC)
• Option 1. The subject’s denial has special standing beyond usual due weight. —SmokeyJoe (talk) 06:13, 20 November 2022 (UTC)

  Why? XOR'easter (talk) 21:14, 3 December 2022 (UTC)

• Option 1. I think weight will depend on case by case and complexity of each situation. A basic and balanced overview of the accusations and of the denials should be the norm. Fad Ariff (talk) 13:18, 25 November 2022 (UTC)

  I don't quite follow the rationale here. If we need to acknowledge that what we write may depend on case by case factors and the complexity of the situation, why endorse a rule that short-circuits judgment? Why would including a subject's self-published denials always and in all circumstances make for a balanced overview? XOR'easter (talk) 21:17, 3 December 2022 (UTC)

• Option 1. If someone who has been accused of something verifiably denies it, we would be remiss to exclude that information. It certainly need not rise to the level of "false balance", but at minimum, "X denies the allegation" should be included. I cannot even conceive of a case where even that brief addendum would violate due weight or create a false balance, so discussing those is unneeded clutter. Seraphimblade ^{Talk to me} 06:39, 3 December 2022 (UTC)

  @Seraphimblade: please explain where you see the word "verifiably" in Option 1. Option 1 says we should always include denials. It says nothing about requiring those denials to be in reliable sources. —David Eppstein (talk) 07:02, 3 December 2022 (UTC)

  Reliable sources depend on context. In this case, "X denies the allegation" would even be reliably sourced to, for example, a self-published source where X denies it. Of course in reality, reliable sources generally will note that an allegation is disputed or denied. Seraphimblade ^{Talk to me} 08:05, 3 December 2022 (UTC)

  "Generally" is not the same as always. Option 1 is a demand that applies in all cases, regardless of what reliable, independent sources say. XOR'easter (talk) 21:19, 3 December 2022 (UTC)

  The proposed sentence would not alter this policy in the way that is being inferred here. This policy has an entire section above this proposed sentence regarding the use of primary sources, verifiability, self-published sources and poorly sourced material. This RfC is in addition to Wikipedia's core content policies, not in opposition of. Homeostasis07 (talk/contributions) 04:27, 4 December 2022 (UTC)

  I've read it, gone away, come back days later and read it again, and I can't see it as a mere "in addition". It's "in opposition". Or, at best, in contradiction with the previous section that you mention. XOR'easter (talk) 16:46, 4 December 2022 (UTC)

  XOReaster, could you expand a bit and explain why you think it is in opposition/contradiction? I don’t think it is, but maybe there is something I am not seeing. Thanks. Blueboar (talk) 17:40, 4 December 2022 (UTC)

  I'm not sure I can express my take more clearly than I have in my earlier comments (probably I can't). The gist is that option 1 demands we do something, whether or not the sources we rely upon indicate that we should. It deprives the community of opportunities to discuss gray areas and edge cases. It insists that the gut-level "gosh we don't want to be unfair" reaction must override careful thinking. XOR'easter (talk) 18:11, 4 December 2022 (UTC)

• Option 3, followed by Option 2 gives people a chance to respond to accusations while preserving the importance of due weight and the argument made by MANDY. Andre🚐 02:38, 9 December 2022 (UTC)

Threaded discussion

Question Do editors feel the addition of the weight and false balance phrases mean that denials shouldn't be included in some cases even if they have been made? This seems to be the core difference between those who support and those who object to the added phrasing. Springee (talk) 12:40, 9 November 2022 (UTC)

• Arguably - and that is a problem. The whole Mandy issue has provoked extensive controversy, with many editors vigorously insisting that denials need not always be included. If 2 or 3 are adopted, I think it is inevitable that some of those editors will argue against including a denial based on those weight and false balance phrases. Otherwise, they will say, what are those phrases there for? That in turn will waste a lot of editorial time and energy. Better wording might to say that a living person's denial of an accusation should always be included, however the extent to which the accused's denial is covered should be governed by weight, etc. In my opinion, the "however" clause is superfluous, but if some feel reference to those guidelines is needed, it should be made clear they apply only to the extent of coverage, not the simple mention of a denial. --agr (talk) 13:26, 9 November 2022 (UTC)

  This captures my thoughts. WP's neutrality policy means that unless a dispute had reached a legal or academic/scientific conclusion (as in the case of pseudoscience which can be disproven), the dispute should be considered unresolved and we should strive to briefly lay out the argument from each side of it, which in the case of a person accused of something that would be their denial. That doesn't mean we need to provide equal time coverage of each side, that should be decided by the weight of sources, but the brief summary should be present as long as some acceptable sourcing (including BLPSPS) can be used for WP:V. If a BLP has been accused but no published source gives their denial, we are not required to include anything about that. Masem (t) 13:33, 9 November 2022 (UTC)

Incorrect wording I'd appreciate if supporters of option 3 would explain why they think WP:FALSEBALANCE is applicable. People seem to be confusing the colloquial use of "false balance" with the actual Wikipedia policy (which option 3 links to). See arguments here on why some of us think WP:FALSEBALANCE doesn't apply at all, or is being incorrectly applied colloquially when it already has a Wikipedia-specific meaning. DFlhb (talk) 12:55, 9 November 2022 (UTC)

I have replied to this in the section above, where I believe it is more on topic. Newimpartial (talk) 17:05, 9 November 2022 (UTC)

Incorrect wording - There's a problem with option 3. WP:FALSEBALANCE is the critical issue, but its test of all sources covering the subject is going to be misunderstood to mean all sources covering the denial. That's a kind of forking. In the context of our WP:NPOV page, it's clear that we are to follow the weight of all sources covering the broader context of the people and events surrounding the denial. But we know from years of stalemated talk page disputes that the language in option 3 leads some editors, usually an enthusiastic minority, to select contemporaneous sources that quote a denial and disregard a much larger body of sourcing that provides the context and reaction to the denial. Perhaps some editors will respond to the question raised above by [[Ping|Springee)) and we can arrive at unambiguous language for the proposal. Otherwise, my current view is that option 4 is the best of the list. SPECIFICO talk 13:41, 9 November 2022 (UTC)

• Answering the original question, yes - my impression is that inclusion of the weight and false balance wording does allow for denials to be excluded. IMO the subject's response to allegations must always be present if the the allegations are. Even if it's a weasel-worded denial that has been widely and repeatedly ridiculed by other parties, even if the evidence for the allegations is incontrovertible, even if the denial doesn't make any sense, even if the allegations are subsequently proven true in a court of law, we must include that the subject does (or did) deny the allegations. Balance and weight determine how we present the response, and how much space we give it relative to the allegations and any responses to the response, but not to the inclusion itself. Thryduulf (talk) 13:58, 9 November 2022 (UTC)

Is this concerning any allegations? GoodDay (talk) 18:42, 9 November 2022 (UTC)

Workshopping a new RFC

I'll note there is as yet no consensus to close the above RFC, or start a new one; I'm just workshopping ideas in case other editors agree a new RFC is warranted. Here's an initial proposal, and please feel free to contribute:

A new RFC, with subsections addressing specific points of contention about denials. There would be a survey & discussion within each subsection, we'd gather the consensus for each subsection, and at the end we'll workshop how to word our changes, and discuss those options. Here are initial subsection headings I can think of:

1. Should policy recommend, forbid, or be neutral on the inclusion of denials for contentious labels? (e.g. transphobic, cult, fundamentalist...) edited 22:14, 14 November 2022 (UTC)
2. Should policy recommend, forbid, or be neutral on the inclusion of denials for allegations of crime, or personal misconduct?
3. Should policy recommend, forbid, or be neutral on the use of self-published denials, for either or both of the above?
4. What should the threshold be, for inclusion of allegations of crime or personal misconduct?
5. If most scholarly sources use a contentious label, should it be stated in Wikivoice, or attributed? updated 22:14, 14 November 2022 (UTC)
6. If scholarly sources haven't weighed in, but most WP:NEWSORGs use a contentious label, should it be stated in Wikivoice, or attributed? edited 22:14, 14 November 2022 (UTC)

I think any discussion of the standard for inclusion of contentious labels belongs in a WP:DUE RFC, not a BLP RFC, but I can easily see the opposing case. "Recommend, forbid, or be neutral" would allow discussions to address these points substantively; feel free to propose alternate wording. DFlhb (talk) 18:49, 9 November 2022 (UTC)

I think this also needs to make sure we go into how contentious labels are used, and why denials might be important. I have seen arguments posited that because no reliable source explicitly denies the contentious label, the contentious label should be treated as a fact statement and written in wiki voice as opposed to being attributed in LABEL. I think this is especially pertinent to labels that implicitly include moral judgements. --Kyohyi (talk) 19:00, 9 November 2022 (UTC)

Agreed, that seems like a relevant issue. DFlhb (talk) 02:15, 14 November 2022 (UTC)

I think another question worth considering is whether the recommendation to include denials/rebuttals is specific to situations where evidence is limited (such as initial accusations and during a trial or hearing), or whether the same standard continues to apply as academic and other high-quality sources accumulate after the events in question. Newimpartial (talk) 19:03, 9 November 2022 (UTC)

I also think the question of my previous RfC, below - whether the recommendation to include rebuttals includes all self-published ones as well as ones that receive RS attention - ought to be considered at an early stage, not after language has already been drawn up. !voters in the prior RfC seem to have held differing views on this that did not correspond to their !votes. Newimpartial (talk) 19:06, 9 November 2022 (UTC)

To be clear, the RFC would be multipart, and all questions would be run concurrently (I've seen precedents of that); it would keep each discussion more streamlined and focused on one point. So the self-published denials question would come before we draw up any wording at all. DFlhb (talk) 02:38, 14 November 2022 (UTC)

I'd choose both, the current RFC and a new one. As I see it, the current RFC is about undoing this WP:BOLD edit from Feb 2021, in light of the objection (see the page history) that it now has WP:SILENTCONSENSUS. I think that should be allowed to run (frankly, I think the edit should be reverted because it was bold, and should be reverted unless it has affirmative consensus at this point, but we can argue about that later).

Separate and apart from undoing that edit, it's worth having a broader discussion as outlined above, which affects multiple policy areas, such as WP:BLPPUBLIC, WP:SPS, and MOS:LABEL. I think those are unclear and would benefit from being clarified. But in the meantime, that bold edit should be undone. Levivich (talk) 19:10, 9 November 2022 (UTC)

I would agree in the abstract, but that revert would never be accepted, and it's simply more productive to focus on this proposed RFC (which, once closed, would remove the entire denials sentence, and replace it with whatever the new consensus supports). For the same reason, keeping the first RFC open to determine consensus on the wording would be futile, when the dispute is about deeper problems of how the policies apply, rather than how this is worded. DFlhb (talk) 19:39, 9 November 2022 (UTC)

I think there are two main issues here. Number one is "what is an allegation?" and number two is "when a subject denies an allegation, what standards should we use to decide whether and how to include that denial?". I don't think those are RFC-ready wordings since they both include a whole lot of sub-issues, but, for instance, we saw a bunch of confusion up above whether an accusation of a crime, a conviction for a crime, an accusation of non-criminal misconduct and a contentious label are all equally "allegations" (and also whether the level of sourcing matters).

I don't like the way you've split the headings above because I think you're mixing issues. What policy should do depends on what an allegation is, and the headings above dance around that question. (I also think that "crime" is a special category here in the minds of many editors and maybe should be split out to WP:BLPCRIME rather than being lumped with "personal misconduct".)

My preference would be to have a two-part RFC. Part one would be Which of the following should be counted as an "allegation" for the purpose of WP:PUBLICFIGURE? and part two would be Given that we are dealing with an allegation against a public figure, which of the following possible denials should be included? Then we can figure out exact wording later. Loki (talk) 19:12, 9 November 2022 (UTC)

Appreciate this. I've tried to stay "high-level" when it comes to grouping allegations, so the policy can stay focused on the "spirit of the rule". That's why I group crimes and personal misconduct together (I think common sense is enough to define the latter: allegations of cheating, and such. We can also amend the examples if anyone wants to, but they're quite clear as-is). "Crimes & personal misconduct" and "contentious labels" seemed to be the two coherent groupings that arose out of the above RFC.

I think asking editors to determine the scope of WP:PUBLICFIGURE would be far too narrow a focus. We should be able to address crimes & labels separately, but your proposal would keep them lumped together, resulting in a constrained discussion.

• If an editor opposes denials for contentious labels, wouldn't they want those to count under PUBLICFIGURE, hanging their hopes on the "due weight and false balance" clause being kept?
• If an editor supports denials for contentious labels, they'd similarly want them to count under PUBLICFIGURE, but would hang their hopes on the "due weight & false balance" clause being removed.

We'd be at an impasse again, since there would be "gaming" of the votes.

For the purpose of the discussion, we should be as open-ended as possible, and focus on how we think the policy should apply, in the abstract; we can figure out how to implement that consensus afterwards. DFlhb (talk) 19:35, 9 November 2022 (UTC)

I see two root objections here:

We should be able to address crimes & labels separately, but your proposal would keep them lumped together -> I agree we should, but I disagree that my proposal does, because those can be separate options under question 2.

For the purpose of the discussion, we should be as open-ended as possible, and focus on how we think the policy should apply, in the abstract; we can figure out how to implement that consensus afterwards -> I agree, which is the whole reason I'm proposing the two questions in the first place. Loki (talk) 00:00, 11 November 2022 (UTC)

I've added two new options, on how we should treat value-laden labels. I feel like having an RFC-based consensus on this would allow us to address this question specifically in the policy, which would obviate tons of talk page disputes. How we treat 'accusations' might also affect how people want to treat denials, so it seems relevant. I think we should launch the new RFC soon, so interested in comments DFlhb (talk) 02:42, 14 November 2022 (UTC)

I notice that in [3], question 5 has been changed from

"Should policy recommend, forbid, or be neutral on the inclusion of denials of contentious labels? (e.g. transphobic, racist, alt-right)" to

"Should policy recommend, forbid, or be neutral on the inclusion of denials for value-laden labels? (e.g. transphobic, cult, fundamentalist...)"

I think the change blurs the issue. "Philanthropist" could be considered a value-laden label. What is at stake here are labels that the person labeled would almost always object to. Such labels are usually applied by people who disagree with the person labeled. Perhaps a better name is attack labels. There is no formal body for adjudicating such labels, as opposed to courts of law for criminal offenses, and they are therefore an expression of opinion rather than fact. As such, BLP policy prohibits such labels from being applied in living persons in Wikipedia's voice. Articles can still cite reliable sources for a subjects views, as well as notable criticism of those views attributed to the critics.--agr (talk) 18:30, 14 November 2022 (UTC)

The problem I see here is the general lack of agreement on-wiki about what labels count as "contentious" (or "usually applied by people who disagree with the person labeled", if that is your preferred operational definition). There has been a high degree of consistency in discussions on BLP pages supporting the inclusion of the labels "far right" and "alt right", for example, even when the BLP subject contests the label, so long as they are widely used in the editorial voice of sources and are not contested in other RS. I for one regard "far right" and "alt right" to be essentially descriptive, rather than normative, just as I regard the term "white supremacist" (when properly supported by sources) to be the correct (objective and neutral) term rather than "race realist". That doesn't mean individual editors or groups of editors - or readers - don't carry value-laden attitudes towards those labels, but the same is true of BLP attributes like "British" or "female" or "athlete". Just because editors and readers carry affect towards a descriptor does not prohibit its use, and BLP subjects do not have to agree with a descriptor to allow its use - WP:BLP has never enjoined otherwise, in spite of some sloppy language in LABEL. Newimpartial (talk) 18:47, 14 November 2022 (UTC)

My edit was just meant to adopt MOS:LABEL's wording verbatim: "value-laden labels", since the debates about denials repeatedly involved MOS:LABEL, and it seems relevant.

I think "attack label" is too subjective, because many people will argue the labels are descriptive, not biased. "Value-laden" seems like the best compromise, since again, MOS:LABEL itself is contested here, and that's the wording it uses. DFlhb (talk) 19:15, 14 November 2022 (UTC)

I think it's worthwhile to point out that while the section header for LABEL is contentious labels. The policy text starts out with Value-Laden labels with a link to our article on loaded language. Certain terms include connotations that include norms and values. Such as things that have the suffix phobic have the connotation of irrational and excessive fear. However what is irrational and excessive is itself normative, what is irrational and excessive only exists within a set of value boundaries. --Kyohyi (talk) 19:28, 14 November 2022 (UTC)

I'm not sure what the thrust of Kyohyi (talk · contribs) above comment is, but indeed LABEL is a short cut to the WP:Manual of Style/Words to watch section "Contentious labels". Furthermore, LABEL applies to all articles. BLP sets a higher standard and the term Contentious label is used there as well. I thought the point of this workshop was to clarify the issues. Changing terminology from that used in the policies under discussion does not seem a step in the right direction.--agr (talk) 21:52, 14 November 2022 (UTC)

That's fair; I missed that the heading for MOS:LABEL was indded "contentious", not "value-laden". I'll change it back, seems proper. DFlhb (talk) 22:13, 14 November 2022 (UTC)

Thanks.--agr (talk) 17:30, 16 November 2022 (UTC)

My point was that contentious labels is largely talking about loaded language. Language that people can argue has a perfectly neutral and descriptive denotation which can also have a connotation that is emotively loaded. I used the example phobic because transphobic is one of the examples listed above. We can have a neutral and even situationally correct denotation on a word, but it can still have a normative connotation associated with it. --Kyohyi (talk) 00:40, 15 November 2022 (UTC)

Ahh, agreed. Another problem with labels like ...-phobic or ...-denier is that they tend to be elastic, often covering far more that the plain meaning of the phrase. So our article on Climate change denial says "Climate change denial can also be implicit when individuals or social groups accept the science but fail to come to terms with it or to translate their acceptance into action." I would propose that any new RFC include an option to prohibit applying a contentious label to a living person in Wikipedia's voice.--agr (talk) 17:30, 16 November 2022 (UTC)