Wikipedia talk:Canvassing

APPNOTE vs INAPPNOTE
APPNOTE has sometimes been interpreted to overrule INAPPNOTE; if a notification is allowed by APPNOTE, then whether it violates INAPPNOTE is not relevant. I believe this is a misinterpretation, and we should clarify APPNOTE to state this, by changing Do not send notices to too many users, and do not send messages to users who have asked not to receive them. to Do not send notices that violate WP:INAPPNOTE, and do not send messages to users who have asked not to receive them. BilledMammal (talk) 15:29, 2 May 2022 (UTC)
 * If nothing else, I think it is much better to make the proposed change, because to too many users is such vague and subjective language that it really is not useful. I note that the edit was made here:, then later reverted while apparently unaware of this talk section: , and I have reinstated it: . --Tryptofish (talk) 16:48, 14 February 2023 (UTC)
 * There was never a consensus to make this change. Rlendog (talk) 18:03, 14 February 2023 (UTC)
 * Well, now it's been changed to another wording: . I'd like to better understand what your objections to it are, since not every guideline edit requires prior consensus and this does not seem particularly substantive, unless I'm missing something. --Tryptofish (talk) 18:09, 14 February 2023 (UTC)
 * It has been almost a year; it is now the status quo and a consensus is required to revert it. BilledMammal (talk) 23:51, 14 February 2023 (UTC)
 * Ah, so the deletionists can make changes without consensus, and then when others want to revert it, they have to get consensus for the change! Makes (no) sense. BeanieFan11 (talk) 15:40, 15 February 2023 (UTC)
 * Tryptofish, the issue is that because of that change, several users (mainly me) were being accused and attacked for "canvassing" for clearly acceptable notifications such as for WikiProjects. BeanieFan11 (talk) 15:46, 15 February 2023 (UTC)
 * OK, the possibility that the change might have resulted in some misuse of the guideline is a valid reason for reconsidering the change, depending on the facts. On the other hand, simply saying that there should have been more discussion a year ago is a poor reason. I'm unfamiliar with the dispute in question, and it's completely unclear to me what the problem with the guideline language was. I'd like someone to explain, specifically, what the problem was, how it resulted from the change in language, and why the previous language was better from that perspective. --Tryptofish (talk) 17:54, 15 February 2023 (UTC)
 * I'm not sure exactly how to put this - I guess that the changed version with the link to the section below seems less clear, and almost circular. In my mind, it makes more sense to describe how a notification that would otherwise be appropriate might not be, as the prior version did, rather than linking to the section below and sort of letting users guess. Hatman31 (talk) 16:32, 16 February 2023 (UTC)
 * Thanks for the explanation. As I see it, however, the to too many users language in the original version suffers from a very similar problem, in that it fails to define "too many". --Tryptofish (talk) 20:30, 16 February 2023 (UTC)

Here are the three versions, side-by-side, for comparison: Is there a way we can reconcile the concerns about these various options? --Tryptofish (talk) 20:41, 16 February 2023 (UTC)
 * Regarding my edit: it was just to avoid using a shortcut as jargon within the prose, which seemed doubly inappropriate given that it just pointed to a later section on the same page.
 * Regarding the original proposed change: sending too many notices is one of the categories described in the section on inappropriate notification. Thus I think the original instruction of not sending too many notices is still covered. But even if the sentence was simply changed to "Do not send messages to users who have asked not to receive them," nothing would change overall regarding the guidance. The following section, "Inappropriate notification" would still be applicable. Perhaps it's best to just shorten the sentence in this manner. isaacl (talk) 22:28, 16 February 2023 (UTC)
 * I like the idea of shortening the sentence. If we do that, I think we should also move that sentence to the end of the next paragraph (for logical flow). --Tryptofish (talk) 23:20, 16 February 2023 (UTC)
 * I think the current paragraph where the sentence is located is about selecting the audience, and the next paragraph is about writing and sending the notice. Thus I suggest keeping the sentence in its current paragraph. isaacl (talk) 01:28, 17 February 2023 (UTC)
 * It would still be applicable, but it isn't always seen that way which is why I originally made this change. For example, Editors who have asked to be kept informed has been interpreted as allowing canvassing of editors who have asked to be canvassed - "Please notify me of all contentious formal discussions related to a broad topic area". BilledMammal (talk) 10:20, 17 February 2023 (UTC)
 * That specific example is an interesting grey area of intersection, as it's common practice for editors to request to be notified about certain discussions. Typically this is for targeted discussions, such as the notification list the arbitration committee set up for discussions on changing the discretionary sanctions system. Perhaps there should be a section on how to respond to such requests: ask the requestor to use one of the page-watching methods described in the "Appropriate notification" section (plus watching the article alerts page for relevant WikiProjects), or in some special cases such as discussions with a specific focus and timeframe, set up a mass mailing recipient list on which anyone can register themselves. isaacl (talk) 23:16, 17 February 2023 (UTC)
 * I think we should have a general review of canvass at WP:VPI and WP:VPR; it will allow us to determine when the community considers it appropriate to notify a partisan audience as well as consider other changes, such as adding a requirement to inform the discussion of any notifications made. BilledMammal (talk) 08:05, 19 February 2023 (UTC)
 * It occurs to me that this may be a case of a double negative not always equaling a positive. Here is what I mean. The actual language in the guideline is: "do not send messages to users who have asked not to receive them." Some editors may misinterpret that to mean that it's always good to send messages to users who have asked for them, but the language that we are discussing here actually does not say that. The language does not tell users to send messages, but rather, tells them not to send messages of a specific type. As far as that goes, I don't see a problem with it. --Tryptofish (talk) 17:08, 19 February 2023 (UTC)
 * I don't understand why that clause is even in this guidance. When has unsolicited individual notification ever been a problem, and how would it not be solved immediately through various other behavioral guidelines? JoelleJay (talk) 21:37, 19 February 2023 (UTC)
 * Although I do not remember details, I do believe there have been cases where editors insisted on notifying uninterested editors, thus annoying them. I think it's reasonable to include as guidance describing appropriate notifications that if someone says they aren't interested in a topic, editors should respect this preference and not, as a general rule, override it with their own personal judgement. isaacl (talk) 22:03, 19 February 2023 (UTC)
 * But it's not like appnote is legislating that individuals meeting the criteria must be contacted, so if someone asks not to be notified why would anyone a) continue doing so due to b) the guideline not explicitly telling them to stop? JoelleJay (talk) 00:20, 20 February 2023 (UTC)
 * The thing is that English Wikipedia guidance isn't written like laws, with specific lines drawn, but as guidance covering intent. This is out of practicality to avoid inordinately long guidance pages (just look at actual laws to see how much fine detail they have to get into), and because so many varied situations crop up that it's not possible to cover all of them. Guidance is generally written assuming most editors are aligned in following a common set of goals, and thus tries to avoid being onerous for these editors. When working collaboratively in a team, it's not unusual for one member to ask another to let them know when certain topics of interest are discussed, and this isn't a problem when everyone is in agreement on objectives. The way to reduce the impact of people showing up who are unwilling to weigh the relative value of the arguments in a discussion is to change the decision-making process such that strength of of an argument isn't determined by the number of people supporting it (or at least not entirely by this). Do that, and notifications of editors with pre-determined positions will be irrelevant. isaacl (talk) 21:55, 19 February 2023 (UTC)
 * Regarding the original expressed concern and current discussions on appropriate notification: it's often a matter of judgement regarding what to do when there is an intersection between the two sections on appropriate and inappropriate notification. I do feel, though, that in the specific case of topic-area related WikiProjects, there is consensus agreement that a neutral notification of the appropriate WikiProjects is appropriate, given that the editors most interested in a topic ought to be involved in discussions on that topic. I appreciate there may be disagreement on whether or not a message at WikiProject talk pages on top of the article alert system is appropriate. (Personally, I think the article alert system is good enough.) Nonetheless, it's relatively common for editors to post notices at WikiProject talk pages, so I don't feel there is consensus against this. isaacl (talk) 22:28, 16 February 2023 (UTC)

I just remembered this discussion, and realized that a lot of time has gone by with no further comment. So I made this edit:, which I think is a minimal solution that does not preclude anything more ambitious. --Tryptofish (talk) 22:29, 5 March 2023 (UTC)
 * I've undone that edit; I believe the clarification that INAPPNOTE applies to APPNOTE is required. BilledMammal (talk) 02:49, 6 March 2023 (UTC)
 * I've come up with an entirely different approach that I think is more suitable. I'd like to invite editors to check the recent changes I made and give me their thougts on it. Thanks. Huggums537 (talk) 06:45, 6 March 2023 (UTC)
 * I also think we should move the sentence Do not use a bot to send messages to multiple pages. to the correct section as well... Huggums537 (talk) 06:53, 6 March 2023 (UTC)
 * I've undone the change; the issue is that some people interpret INAPPNOTE as not applying to APPNOTE, so we need working in APPNOTE that INAPPNOTE applies. Lets discuss this and get a consensus for any change before editing the page. BilledMammal (talk) 10:43, 6 March 2023 (UTC)
 * That's OK with me, but consensus is not simply letting discussion go silent and then reverting as soon as someone makes an edit that reflects (I think) what was discussed last. --Tryptofish (talk) 19:35, 6 March 2023 (UTC)
 * On that topic, I stand by my earlier suggestion which received no response: I think we should have a general review of canvass at WP:VPI and WP:VPR; it will allow us to determine when the community considers it appropriate to notify a partisan audience as well as consider other changes, such as adding a requirement to inform the discussion of any notifications made. BilledMammal (talk) 20:25, 6 March 2023 (UTC)
 * Do please leave a note here if you start such a review. --Tryptofish (talk) 20:33, 6 March 2023 (UTC)
 * I was hoping that other editors would provide some input on scope and initial topics first. BilledMammal (talk) 20:43, 6 March 2023 (UTC)
 * Oh, I see, that's reasonable. Sorry, I don't have any good ideas, because it hasn't really been a problem that I've experienced. (My concern is mainly that saying "too many", in a version that is no longer current, was too vague to be useful.) --Tryptofish (talk) 20:44, 6 March 2023 (UTC)
 * I won't start it immediately - I've currently got two threads going at VPI/VPR and don't want to add a third yet - so hopefully someone else will see this discussion and add their thoughts. BilledMammal (talk) 20:57, 6 March 2023 (UTC)
 * I did respond indirectly. I don't think notification is the heart of the issue. I think we'd be better off figuring out better ways to evaluate the expressed viewpoints, so raw numbers are not used as a proxy for strength of argument. isaacl (talk) 22:26, 6 March 2023 (UTC)
 * I agree, but that is a much harder problem to solve, and until we do we need to handle the notification issue. BilledMammal (talk) 07:52, 7 March 2023 (UTC)


 * comment: Is there any particular reason why there is such an intense focus on mixing these two sections together that we have to have two links to the same section in the same sentence? What's with all the overkill?!? Odd. Very odd. Huggums537 (talk) 03:46, 7 March 2023 (UTC)
 * I'm using the wording introduced in . I don't have a preference between this wording and the former wording. BilledMammal (talk) 07:52, 7 March 2023 (UTC)
 * I agree that the double-linking was a problem, so I made this edit: . --Tryptofish (talk) 21:06, 7 March 2023 (UTC)

Discussion at WP:VPI
As discussed, I've opened a discussion at VPI about possible modifications to CANVASS. BilledMammal (talk) 07:53, 18 April 2023 (UTC)

Minor point of order
Above, I see a discussion about whether WP:APPNOTE should explicitly bar activities discussed in WP:INAPPNOTE. Prior to this discussion, there was no such explicit bar. In other words, the status quo was that APPNOTE should not mention INAPPNOTE. Then, a discussion was started at WP:VPI, and the post there included the question, "Should the examples listed at WP:APPNOTE be exceptions to WP:INAPPNOTE, or just examples of notifications that are usually acceptable?".

I don't see a clear consensus—either on this page or on the VPI page. Shouldn't this guideline maintain the status quo until such a consensus is established?--Jerome Frank Disciple (talk) 18:33, 30 April 2023 (UTC)
 * The current version is that status quo; it has been in place for a year. It shouldn't change without a consensus to do so. BilledMammal (talk) 19:02, 30 April 2023 (UTC)

Status quo in lead section
Back in January, there was an undiscussed bold edit added to the lead. I recently found out about it, and reverted it with substantial reason for disagreement. This should have been enough to keep controversial material out of guidance until it has been further discussed, but my removal of the objectionable edit was reverted with an explanation that, [in spite of any objections to it], the edit is still useful and "right" considering the immediate context. I reverted the bold edit again with a counter argument that the context also strongly suggests otherwise, and a reminder that bold edits to guidance should be taken to discussion if they are controversial before they are added. I have a strong opinion that "special use" for words on Wikipedia that would otherwise have a normal meaning is an abuse of language becoming like a spreading cancer. It seems we are just making WP:CREEPy stuff up just to justify other CREEPy stuff. Huggums537 (talk) 06:24, 2 May 2023 (UTC)
 * It seems to me that we each explained our reasoning in our respective edit summaries. Here is a link to the Wiktionary definition of the term: . The way that we use the term on Wikipedia is pretty obviously different than that. The closest is the definition that comes last (as the least frequently used), and is presented as a political use of the word. As such, it seems reasonable to me to explain what we mean by it. If your objection is to the use of "special words", that would be addressed by renaming the guideline page, not by removing all explanation of how the use is distinctive to Wikipedia, as a term of art. And we use such terms, in WP: space, for a lot of other things beside canvassing, so I think that your objection may be a matter of personal opinion, rather than community consensus. --Tryptofish (talk) 17:28, 2 May 2023 (UTC)
 * I think it is the addition itself that is the matter of controversy regarding personal opinion rather than community consensus since there has been no policy or guideline offered that supports such a consensus, but only personal opinions related to personal interpretations of Wiktionary definitions that really make no sense whatsoever since even if we say the least frequently used most lowly definition on Wiktionary is equivalent to any use on Wikipedia, then we would be wrong for saying Wikipedia use compared to Wiktionary use are "obviously different", and there would be plenty of room for debate, or differing opinions without any actual guidance or policy with a real true consensus to look to. For another thing, if I can make the removal by using definitions just as easily as you can argue that the addition should be there with your interpretations of Wiktionary, then it is obviously evident the addition is based on opinions rather than policy no matter which side of the fence you land on so it isn't fit for guidance. This is the crux of my objection, that the addition was added without any support or backing from policy or guidance, and then no discussion from community for consensus. My opinion about "special use" words was just an added footnote, but I'm glad you tried to latch onto it anyway. Huggums537 (talk) 22:08, 2 May 2023 (UTC)
 * Besides, the addition is just plain wrong anyway. It states that the Wiktionary meaning does not have any negative connotations, and that is simply not true at all as I tried to explain in my edit summary from the original removal. Looking closely at the Wiktionary definition we can use the political term that you wanted to use which is #4 where they use the word "solicitation", and if you check the Wiktionary definition on that you see that #2 has some very negative connotations, but going back to the definition of canvass if you check #3 (which BTW could also just as easily apply to Wikipedia as #4) you will see that to seek or solicit opinions or support from people contains that word again with negative connotations you can check on Wiktionary #3. Huggums537 (talk) 22:39, 2 May 2023 (UTC)
 * I guess to be more concise, we should leave out any additions that might be false, misleading, or contain technical errors of any kind. To suggest that the real world use of the word "canvass" doesn't ever have any negative connotations whatsoever is really just absurd when you think about it. Huggums537 (talk) 04:54, 3 May 2023 (UTC)

Notifying editors for follow-up AfDs
I was wondering if it is okay or even encouraged to notify editors who were involved in the 3 previous AfDs (#1; #2; #3) about a follow-up AfD. They helped making the decisions in the previous AfDs, and perhaps should be notified of a new nomination that is a direct follow-up of those previous AfDs. WP:APPNOTE says Editors who have participated in previous discussions on the same topic (or closely related topics) may be notified on their user talk pages (so not the AfD page itself), and WP:VOTESTACKING similarly states: Posting an appropriate notice on users' talk pages in order to inform editors on all "sides" of a debate (e.g., everyone who participated in a previous deletion debate on a given subject) may be appropriate under certain circumstances.

However, I am wary that this could still potentially violate WP:VOTESTACKING. After all, I know how they voted and commented in the past 3 AfDs (all voted either Delete or Merge), so I can guess how they are going to vote and comment in a follow-up AfD, thus perhaps unduly influencing the outcome of the discussion. If the voting had been more mixed, I would definitely have notified everyone regardless of their votes. But in practice, notifying them now seems like votestacking, even if it would technically be allowed on their user talk pages. I don't want to unduly influence the result, and think the nomination should be judged on its on merits, even if those are partially dependent on the precedent set by the previous deletion of the other 3 articles. So, shouldn't I notify them? I'm inclined to think it would be inappropriate in this case. I've just never done this before in AfDs, so I better ask first. Cheers, Nederlandse Leeuw (talk) 13:55, 9 June 2023 (UTC)

Remove "Appropriate notification: On the user talk pages of concerned editors"?
I think it is rarely appropriate to summon a specific editor to a content dispute via user talk. The only exception I can think of is if a page is being nominated for deletion, then a user talk notification notifying the original author of the page is appropriate. Instead of using user talk to summon specific people, the ping system can be used to summon a group of people, such as every participant in a previous RFC or ever editor to a page. Pings are more transparent and are more likely to be used to summon a group rather than cherry-picking folks with a POV.

Perhaps it is best to significantly trim down or completely remove "Appropriate notification: On the user talk pages of concerned editors" and its sub-bullets, with the goal of not encouraging or legitimizing this behavior. Thoughts on removing this? – Novem Linguae (talk) 05:23, 4 October 2023 (UTC)
 * I support this; notifications should be transparent, and when talk page notifications are used too widely they can become spam, as we have seen in a few recent cases. Exceptions will exist for things like ARBCOM elections, but I don't think such notices are covered by this policy anyway. BilledMammal (talk) 05:40, 4 October 2023 (UTC)
 * Numbering them, as far as I can tell 1 and 3 would seem to fall under inappropriate canvassing, while 2 and 4 would not. - jc37 08:22, 4 October 2023 (UTC)
 * The notification system formerly known as Echo is opt-out, and thus community consensus so far is that individual notifications should be done on user talk pages. If the issue is letting others in the discussion know about notifications that were made, then the community should reach a consensus on guidance for this aspect. isaacl (talk) 16:58, 4 October 2023 (UTC)
 * Another issue with pings is that they don't necessarily work all the time, as when there isn't a signature in the same edit as the ping. I don't think we should change anything to say that pings are preferable to talk page messages.
 * About simply deleting the material, I think that the fourth bullet point, "Editors who have asked to be kept informed", has a unique status as being the one thing where editors agree that it is appropriate, and is not canvassing. So instead of a complete deletion, I'd suggest trimming it to one line, without the subsequent bullet points:
 * "On the user talk pages of editors who have asked to be kept informed."
 * Deleting the rest doesn't change those things from appropriate to inappropriate, but just leaves them out, and it's true that there are constant disputes about them, that we could better do without. But by implying that user talk messages to editors who haven't actually asked to be informed is not always appropriate, I think this would better align the language with current practice. --Tryptofish (talk) 19:17, 4 October 2023 (UTC)
 * #2 is also fairly common practice. You see it most on things like XfD or RM, when notifying those who commented in a previous discussion, or on a noticeboard post about a similar topic under discussion. - jc37 20:29, 4 October 2023 (UTC)
 * Yes, it is, but I've also seen numerous disputes about whether it's appropriate. That's why I said "Deleting the rest doesn't change those things from appropriate to inappropriate". --Tryptofish (talk) 20:33, 4 October 2023 (UTC)
 * YMMV, of course, but the only disputes over it that I recall seeing are when they didn't ping everyone, or in some other way (intentionally or unintentionally) cherry-picked pinging some but not all. (which, of course, would fall under vote-stacking section.) - jc37 20:59, 4 October 2023 (UTC)
 * As it happens, from my watchlist today: . (Not that I want to canvass anyone to go to that discussion! {) This is where an editor apparently did contact everyone, but two other experienced editors had concerns about it. --Tryptofish (talk) 21:19, 4 October 2023 (UTC)
 * Wow. I can't begin to list the number of issues I see in that thread.
 * But to me, that's all the more reason that it should be laid out here clearly. - jc37 21:34, 4 October 2023 (UTC)
 * Ha! Maybe that's something that fits better with a right-versus-wrong way presentation, for laying it out clearly, instead of just listing it as "appropriate". --Tryptofish (talk) 21:39, 4 October 2023 (UTC)
 * Maybe. But my concern is that there are lots of applications of this, and we'd just be opening the doors to more wikilawyering... - jc37 22:08, 4 October 2023 (UTC)

A related WP:GAMING proposal
Followers of this page may be interested in Wikipedia talk:Gaming the system. — SMcCandlish ☏ ¢ 😼  08:39, 1 November 2023 (UTC)

More strict guidelines for what an editor should do when the subject of canvassing
Most of WP:CANVASS deals with discouraging users from canvassing (understandably.) It briefly mentions how to respond to canvassing, but (implicitly) from the perspective of a bystander who sees it happen, not someone who is the subject of it. This recent ArbCom motion makes it clear why that's not sufficient - banned editors are canvassing users via email; guidelines that only target the canvasser are obviously going to be ineffective in that situation. And, as discussion there makes clear, stealth / email canvassing has been occurring with increasing frequency. I think that the initial reaction of ArbCom in that case reflects the general community consensus and actual practice; however, this page doesn't actually lay it out that I can see. So I suggest adding a "what to do if you are canvassed directly" section or something along those lines, perhaps in WP:STEALTH, stating that: The first point is, I think, necessary because if an editor doesn't act on the canvassing, they themselves haven't done anything wrong; we'd prefer they report it but we can't realistically require that. The second point is necessary in order to discourage stealth canvassing by making it more difficult. The parenthetical is necessary because I believe the editors who are most likely to report being stealth-canvassed are highly experienced and extremely active ones, who are the very ones who have the strongest claim to saying "ah, I'd have seen the discussion on RFC/All or AFD anyway, so I wasn't really canvassed"; it's important to make the requirement to report being stealth-canvassed as clear-cut as possible. If an editor believes they would have participated anyway, that's fine, they can say so when indicating they were canvassed; but they still need to make it clear so anyone closing the discussion or evaluating its consensus can make their own call on that and in order to ensure that stealth canvassing remains difficult. --Aquillion (talk) 07:18, 5 January 2024 (UTC)
 * You are strongly encouraged to report it; but as long as you don't act on it, you aren't required to.
 * However, if you do participate in the discussion you were canvassed for (even if you believe you would have participated anyway), you are required to disclose that you were canvassed. Failure to do so may lead to sanctions.
 * Thanks for thinking of this, and I support such a change. One revision to your proposal: on the first bullet point, I would change "; but as long as you don't act on it, you aren't required to" to ", regardless of whether or not you act on it". Although I agree with the concept you describe, that we shouldn't require it to the extent of making non-reporting sanctionable, the language seems to me to go too far in implying "don't worry about it". As long as we say "strongly encouraged", that makes it clear that we aren't saying "required". --Tryptofish (talk) 18:42, 5 January 2024 (UTC)
 * I think there's a gap between the two bullets that needs filling, even signposted with the use of "however" in that second bullet. I agree that if you respond to direct canvassing by taking part in the discussion to which you were summoned, you should disclose that. But it also seems that if a good-faith editor receives an invitation that they recognize as inappropriate canvassing, whether they choose to report it or not, the optimal response is not to participate in the discussion, and we should put that in writing, too. I recall the case of a fairly prolific editor who was in an ANI discussion clearly leaning toward his ban, who reached out to a couple of other editors to come defend him. I was quite impressed that one of the canvassed editors responded that they had been inclined to do so, and had been preparing comments, but that the receipt of the canvassing invitation made their participation no longer appropriate (and even scolded him for putting them in that position). We can't force that sort of response, but it's something we should encourage. In discussions that are significantly less oriented toward any sort of voting, canvassed editors who choose to participate can still effect a significant impact on the tone and course of the discussion, shaping it in a different direction to a very different outcome. Grandpallama (talk) 23:36, 5 March 2024 (UTC)
 * I'm reluctant to go beyond the disclosure requirement because ultimately editors can't control whether someone attempts to canvass them or not; I think there's limits to how much we can suggest they do in response (especially given that even suggestions in policy are often taken to have enough force to lead to sanctions if ignored.) Barring an editor from an entire discussion - possibly a very important one, which they would inevitably have participated in anyway - purely because of something someone else did is too much for me even as a suggestion; the disclosure is IMHO enough in that anyone who eventually assesses consensus will know to take that into account. --Aquillion (talk) 21:00, 2 April 2024 (UTC)
 * I would support this, and actually made a bold edit to that extent two years ago, but unfortunately it was reverted. BilledMammal (talk) 00:27, 1 May 2024 (UTC)