User talk:Robert McClenon/NOTLAB

I'll look at this more tomorrow morning, but I'd say the first two sentences are right out. They're just not necessary. Maybe go for something more like, Wikipedia is not a laboratory for experiments using editors as unwitting participants.

The informed consent piece is a nice one, and one I hadn't previously considered. I also think that ban is jumping the gun a little bit, since a ban requires a community consensus, and I would go for more like blocked pending further discussion, to empower administrators to block first, and then let the community discussion happen when there is ample evidence this is happening, rather than having an extended discussion and blocking afterward. GMG^talk 22:29, 6 October 2017 (UTC)

I think it would add weight if there was consensus that these experiments violate the wmf:Terms of Use, per the section Engaging in False Statements, Impersonation, or Fraud > Engaging in fraud (ie. editing under false pretenses)... and/or they represent a covert form of disruptive editing. If those were accepted, language to that effect could be added. --Drm310 🍁 (talk) 05:20, 7 October 2017 (UTC)

I think there's already pretty widespread agreement that this type of editing is disruptive, and this policy would only codify that consensus already reached on a number of ANI discussions. And for whatever it's worth, I've rarely if ever seen a discussion about the interpretation of the TOU to be anything close to productive. I've given a go at a second rewrite. As I said above, I think the ethical research concerns are really central to the whole thing, and I threw in the threat that we would reach out in the real world to their IRBs just to boot, because... well... we will. GMG^talk 12:56, 7 October 2017 (UTC)

Good second version. I will look at it more shortly. Robert McClenon (talk) 15:26, 7 October 2017 (UTC)

Other than reworking a draft until it's perfect, there are at least two additional outstanding problems:

1. If someone wants to do research but wants to do right and submit it for community review (assuming we keep those bit), then where should they be directed to go to seek consensus? I'm thinking maybe Wikipedia:Village pump (proposals), but I'm definitely open to alternatives.
2. We still need to put together a detailed problem statement, to show why this is needed, and avoid the "problem in search of a solution" issue, that crops up any time anyone proposes anything at all. GMG^talk 16:01, 7 October 2017 (UTC)

I'm also going to take the liberty of pinging User:Masem here for their opinion, as someone who has disagreed with me more often than not. GMG^talk 22:53, 7 October 2017 (UTC)

Third draft

I have put together a third revision, incorporating many of the missing aspects in the second version, which were covered by User:Jytdog in their draft proposal at Wikipedia talk:What Wikipedia is not#drafting. Major changes included the addition of multiple links to view existing research, and explicitly covering behavioral guidelines, COI, and PAID. I moved the links to existing research to a footnote since the list was getting prohibitively long. Much of the content linked to in places like Wikipedia:Research#See also was omitted, because they're either historical, inactive, or otherwise generally stale.

I have also added a "best practice" guideline that hopefully addresses concerns about discouraging legitimate research, and specifying that research is not prohibited per se, but must be done cautiously, depending on the design. Also note that I have intentionally omitted specifying which public forum is the appropriate place to seek approval of intrusive research, since I suspect this may vary depending on the research being done. The village pump comes to mind as a good middle ground, but I expect that exceedingly minor or specific research may be appropriate to bring up at an active WikiProject, and exceedingly broad research may need approval at the highest level, in a place like AN. GMG^talk 12:01, 8 October 2017 (UTC)

The researchers are unlikely to consider themselves as "paid editors" and several people have (incorrectly) chimed in that the editor was wrongly labelled a "paid editor". The way you wrote this leaves it for the research to self-define whether they are a paid editor and whether they have a COI .. i wrote the draft i proposed to define this for them -- " If experimenting editors are paid for their work in Wikipedia they must disclose that per the WP:PAID policy, and if they are not paid they still should disclose any outside interest that is driving their activity in Wikipedia, per the WP:COI guideline."... rather than leaving it for them to define themselves. So many academics come to WP to promote their own work or push their own theories, oblivious as to what that constitutes a COI here in WP. (Wikipedia:Conflicts of interest (medicine) actually addresses this, as does WP:SELFCITE in the COI guideline). Jytdog (talk) 16:42, 8 October 2017 (UTC)

I'm changing the loss of editing privilege piped redirect from WP:BLOCK to WP:BAN, because the latter is a more serious action and is really what is needed with the rare cases such as started this. An editor can get unblocked by saying that they understand. Robert McClenon (talk) 17:10, 8 October 2017 (UTC)

Umm... I mean, I've seen folks say "I don't work for the author, I work for the publisher, so there is no conflict of interest." So, there's only so much idiot-proofing you can do. But what if we strike This includes conflicts that arise from affiliations with academic or other scientific institutions, and editing which is ultimately paid for by things like grants or fellowships. and include If users who are involved in research are paid for their work on Wikipedia they must disclose those relationships per our paid editing policy, and if they are unpaid, they must still disclose any outside interest driving their activity on Wikipedia.?

As to Robert's objection, I think it just... lessens the likelihood of passing in a purely pragmatic sense. Policy can enable admins to block under certain circumstances, but only the community can authorize a ban. That's fairly nuanced, but it's already been brought up once in opposition, and I think it probably just makes it a weaker proposal that we should go to the extreme. Any user who is blocked, but seen to continuously abuse the project, can then be banned, but the only people who can be banned are those already blocked according policy. GMG^talk 23:13, 8 October 2017 (UTC)

Good point. Do you just want ban reverted to block, or would you be satisfied with a statement that such editors may be blocked or even banned? Robert McClenon (talk) 02:23, 9 October 2017 (UTC)

(Sorry for the length.) The biggest place I could see a potential for bans would be a TBAN, but a TBAN from conducting unethical research is a little like a TBAN from randomly inserting the word "fuck" into articles. It's not a restriction; it's just the normal behavioral standard. In a case where we actually needed a TBAN for an otherwise productive editor (which would amount to a "last chance" and empower any admin to block on sight for violation), it would need a detailed community discussion anyway, and would mainly be in lieu of a block, to allow them to productively edit elsewhere. So you can essientially say that "block" is the standard, and the community can always walk that back to a TBAN if warranted as an act of leniency.

That brings us to site bans. Most of the cases where we have trouble have been from folks who come to Wikipedia for the sole purpose of using volunteers as a petri dish. In those cases, the most important thing is that they be technically prevented from causing ongoing damage. In cases like this, they were here for the purpose of publishing, and were, as one editor put it, here to take the community for a ride. They wasted an inordinate amount time, but yet we immediately get the argument that no action should be taken, since they were disrupting "in good faith", and their defense seems to be that they caused as little disruption as was necessary to accomplish their research goals.

One of the main purposes of this proposal IMO is to clarify that that is fundamentally not the standard, that intrusive research is presumed prohibited unless the community explicitly permits it, and that admins should err on the side of protecting volunteer time when something is underfoot. The standard for that protection would be a block, and not a site ban, which effectively just raises the bar for unblocking, from an unblock request and a caffeinated admin, to a community discussion. The idea is to shift the burden of proof and bring it more in line with something like an institutional review board. In other words, the onus is not on us to demonstrate that they're disruptive enough to warrant a block; the burden of proof is on them to demonstrate that their research is done ethically, as it is in any other case of research involving human subjects. GMG^talk 10:47, 9 October 2017 (UTC)

Beneficial content

A point made at WT:NOT is that we don't want NOTLAB to be used to nix projects that contribute content that is others fully within policy and expected inclusion. If the end goal is to get some improved articles that follow the core content content policies, ala the Gibraltar tourism stuff, and see how those propagate outside WP, great! This is not an issue, though we would want to be told that you were doing this as part of a research project, just so admins wouldn't become suspicious of a sudden interest by multiple areas in a narrow target area. --MASEM (t) 15:32, 9 October 2017 (UTC)

I don't think this would apply as intrusive, as a matter of more or less common sense, since the purpose is to improve the encyclopedia, and measure the impact of that improvement. As it is currently written, it is still advised as best practice for these editors to post a user page notification, which as you seem to get to, would avoid something like an accusation of COI, when others come across a very narrow and aggressive editing campaign afoot. GMG^talk 15:50, 9 October 2017 (UTC)

A problem with the Spanish tourism article and with the chem articles that spurred the thread, is that the experimenting editors thought they were adding good content, when in both cases they were adding bad content. The Spanish-tourism people dumped a bunch of unsourced/badly sourced promotional shit into four Wikipedias, and the chemistry people dumped a bunch of WP:NOTJOURNAL content into Wikipedia. Both of these were done for some end other than improving Wikipedia. The chemistry editor basically said "fuck you" to the editing community and blew us off; i have not studied yet how the spanish=tourism people actually interacted with others.

In both cases, volunteer effort was sucked up so that these people could do shitty correlational research. I am sure that neither team took any consideration of the effort of the volunteer community. This is not OK. Jytdog (talk) 21:55, 9 October 2017 (UTC)

Well, User:Jytdog says that the chemistry editor basically said "fuck you" to the editors. She said nothing at all, nothing, nothing, nothing. I can see equating that to a profanity. I still don't understand the experiment, but the failure to respond at all was, as Jytdog says, disrespectful to the editors. Not only did she never respond to the comments, but she continued to submit the same articles repeatedly. That is the most puzzling thing. I agree that it was not okay. I still don't understand what the experiment was meant to do. It is clear that the intent was not to improve the encyclopedia, or at least that there was no intent to improve the encyclopedia in accordance with its own policies, because that would have involved discussion. I still think that the chemistry content was almost suitable for Wikipedia, except that the editor wasn't working with the volunteer editors. Any experiment that doesn't involve communication with the volunteer community is an experiment that is not beneficial to the encyclopedia. Robert McClenon (talk) 23:42, 9 October 2017 (UTC)

As I noted at the discussion at the village pump....(just copy/pasted from there)

I checked the edit count for Carolineneil, and they used a talk page exactly once back in November 2015, here at WT:CHEM, and they wrote These articles aren't class assignments. They're part of a project, with Dario Taraborelli at Wikimedia, to bring more advanced scientific content to Wikipedia. There were extensive discussions with Dario before the creation of these articles. -- User: Carolineneil. That is the definition of NOTHERE not to mention arrogant as hell. If they would have taken a different approach, a bunch of the time of the editing community would not have been wasted, and their contribs would have been more productive. (the whole section that snippet is from, is here)

That single reply and all the non-replies, non-engagement, etc etc is what i was summarizing when I wrote basically said "fuck you" ("basically said", not "said") to the editing community - in other words, they expressed contempt for the editing community. Jytdog (talk) 23:49, 9 October 2017 (UTC)

btw, the point of the experiment was to see if actual language used in Wikipedia was picked up and used by the scientific community. as they wrote: "Incorporating ideas into a Wikipedia article leads to those ideas being used more in the scientific literature." that is what the headlines about "Wikipedia shaping science" are about. (in my view it is a badly controlled correlation study with boatloads of potential confounders) They needed to get the language into WP in order to the do the experiment. That is what the edits were for. Jytdog (talk) 23:54, 9 October 2017 (UTC)

Apparently, I didn't look into the entire thing in as much depth as I should have. I would still think that the "best practice" steps would have made things easier all around. If someone can see that they're being disruptive, and see that they're doing research via a prominent notification (which would be required across the board...maybe that language should be made stronger...more than a suggestion for best practice), any editor who picks up on it can say "this is intrusive enough to need approval", and raise that with the editor. If their response is basically fuck you, then it's a short trip to bring that up to the community with:

1. Is this intrusive enough to need approval?
2. Are they refusing to comply with this?

If the answers to both 1 and 2 are yes, then a block would normally be warranted, and the user can explain in an unblock request how they plan to comply with ethical research expectations going forward. I imagine the conditions of an unblock would likely include a proposal for community approval of the research design, and an unambiguous agreement that they are not going to proceed until they have that consensus.

I don't think we can hope to set a hard fast rule about what is and is not intrusive/ethical. It's always going to have some level of subjectivity, which is why IRBs exist. But ideally, we can empower page watchers and recent changes reviewers to spot bad research when it happens, and make a judgement call without needing to argue at AN/ANI about the basic expectations for researchers, which hopefully we've already spelled out with this exact proposal. GMG^talk 10:28, 10 October 2017 (UTC)

Tweaks

• I have done a number of tweaks to strengthen the wording throughout. As always, comments or complaints welcome. GMG^talk 13:26, 10 October 2017 (UTC)

Too long

If this is for NOT, it needs to be way shorter. Jytdog (talk) 13:51, 10 October 2017 (UTC)

Still way too long if this is intended for NOT. If so basic notions like "the encyclopedia that anyone can edit" don't have to be rehearsed. Jytdog (talk) 14:01, 10 October 2017 (UTC)

I managed to cut out 90 words without, I think, losing any real meaning. Not sure if you saw my revision, or the copy/paste I did initially in order to do tightening in a separate section. GMG^talk 14:11, 10 October 2017 (UTC)

Ok. I've now cut out a total of 107 109 words, or slightly more than a quarter of the size of version three. GMG^talk 14:19, 10 October 2017 (UTC)

Are We Ready?

Are we ready to post this to the Talk of of WP:NOT as an RFC? Robert McClenon (talk) 03:37, 13 October 2017 (UTC)

I was kindof leaving it open to see if there were any follow up objections. I'm fine with it if nobody has anything else to add. Gratuitous pininging of everyone who's commented: @Jytdog:, @Drm310:, @Masem:? GMG^talk 12:11, 13 October 2017 (UTC)

I think it's good to go. It should cover all the bases, from amateur breaching experiments to covert research by academics. --Drm310 🍁 (talk) 14:07, 13 October 2017 (UTC)

I am concerned about the "outside Wikipedia" stuff. We cannot obligate people to do anything outside of WP, so saying they need outside IRB approval is outside the scope... and my sense is that people would object strongly to including "contacting their IRB" in NOT. I also tightened a lot - I don't think this should be more than one paragraph. There was a lot in the prior draft that was great. Jytdog (talk) 15:32, 13 October 2017 (UTC)

Will look over and reply when I can get on an actual keyboard. GMG^talk 16:45, 13 October 2017 (UTC)

• Okay, I've made a number of tweaks. I tried to be clear in my edit summaries so I'll just leave those for review, rather than trying to explain them again here. The one thing that does still give me pause is whether we should include the village pump as an intermediary to WikiProjects and AN. It seems like a natural middle ground, and I can imaging someone objecting based on the expectation that this is going to end up cluttering AN with a whole slew of research reviews. GMG^talk 12:28, 16 October 2017 (UTC)

Do we go with Quatro, or with the subsequent section? Robert McClenon (talk) 16:09, 16 October 2017 (UTC)

I've been editing on the latest section. But I wouldn't be too hasty about starting this. I had considered waiting until the other ongoing RfC was closed. GMG^talk 16:40, 16 October 2017 (UTC)

fixed a couple of things. this is good to go as far as i am concerned and i see no advantage in waiting. i would suggest posting it for further feedback (making that clear so people don't start !voting) before actually launching the rfc. Jytdog (talk) 17:36, 16 October 2017 (UTC)

Maybe it would be best to post a follow up comment to the existing discussion at NOT, rather than post a draft? We can collapse the previous drafts here to avoid confusion, and hopefully pointing them here instead of posting the text there will avoid people starting to fire from the hip with supports and opposes. GMG^talk 17:39, 16 October 2017 (UTC)

sure. Jytdog (talk) 18:16, 16 October 2017 (UTC)

Incidentally, why exactly is the policy shortcut expanding to say "Wikipedia:NOTLAB" instead of displaying "WP:NOTLAB", which is the text that's entered into the policy shortcut template? Does it default to that when the redirect hasn't yet been created? GMG^talk 19:47, 16 October 2017 (UTC)

I'm doing a review. Will edit the last version. Legacypac (talk) 23:10, 16 October 2017 (UTC)

Village pump

I'm just going to start this as a new section, because this is a serious concern of mine regarding the likelihood of this passing. Should we include the village pump as a potential venue for approval? GMG^talk 23:33, 16 October 2017 (UTC)

I don't understand. What is your concern? Are you concerned that the proposal is likely to pass, or that the proposal is not likely to pass, or that experiments are too likely to pass, or that experiments are too unlikely to pass? I want to be nearly sure that the proposal passes, and anything that reduces the likelihood of experiments passing is probably desirable. (Experiments whose purpose is the improvement of the encyclopedia or of access to the encyclopedia are not the issue. They are encouraged. I would like to minimize the likelihood of any other experiment passing.) Please explain. Robert McClenon (talk) 23:13, 17 October 2017 (UTC)

I was anticipating objections like those in the thread immediately below, that the village pump is the happy medium for consensus on pending research. GMG^talk 23:15, 17 October 2017 (UTC)

Consensus-related objection, and suggested fix

This seems entirely wrongheaded to me: "prior consensus at relevant WikiProjects or at the administrators' noticeboard".

1. Wikiprojects have no magical authority of any kind, even over topics they consider within their scope. As a matter of policy (WP:LOCALCONSENSUS) the are nothing but some editors who share an interest in a topic.
2. Admins do not make up the community's mind for us.

The proper venue for such a research proposal is WP:Village pump/Proposals. I agree with all the rest of the wording in the fifth draft ("current" as of this writing).  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  10:40, 17 October 2017 (UTC)

The idea, at least in my mind, is to get it to sunlight somewhere. As is the case with close reviews and standard offer requests, taking it to AN is not necessarily for admins to weigh in as admins per se, but rather to simply get it in the most public place possible to let the community weigh in. Admins who reopen a discussion or unblock someone who's indeffed or community banned, aren't really doing it as a "decision maker" but as a "consensus enact-er". AN just happens to be the most public place to get community input.

Ideally, WikiProject members will have a good enough head on their shoulders to bring a proposal to a more public forum, like VP or AN, if the scope is beyond the project. But the idea of including projects is to provide a "least-public public-forum" to avoid clogging up places like VP or AN with uncontroversial proposals, like testing different user talk welcome templates, as was suggested as an example on the main discussion. Bringing something like that to somewhere like AN would be pro forma to the point of needlessly wasting time, which is what this is intended to avoid in the first place.

At the very least, I'm going to add VP as an option, based on your concerns and the fact that no one has replied to my own concerns above. GMG^talk 10:55, 17 October 2017 (UTC)

Re: "Ideally, WikiProject members will have a good enough head on their shoulders to bring a proposal to a more public forum, like VP or AN" – cut out the assumption and the middleman, and just require the more public forum. AN is not a more public form, it's a noticeboard for administrators which most editors pay no attention to. Re: "if the scope is beyond the [wiki]project" – there is never going to be any such thing as an experiment limited on only a wikiproject because wikiprojects are not fiefdoms; all their content is editable by everyone on the system, and their "membership" is fluid anyway. If someone wants to propose using WP in some kind of experiment or study, of a nature that needs disclosure and community buy-in, the venue for that is WP:Village pump/Proposals. It's named that for a reason. This is looking like "If people want to delete a template, they should take it to TFD, the process for that." "No, they should take it to TFD or AN or some wikiproject." "Um, no, they should take it to TFD, the process for that." This is quite simple, really.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  23:22, 17 October 2017 (UTC)

Any objections? GMG^talk 23:27, 17 October 2017 (UTC)

Merely building an encyclopedia?

Is "building" the encyclopedia the only goal? I'm thinking of initiatives like the WMF's Wikipedia Zero or other projects whose goal is to expand the reach of already existing content and improve the ways that the content is used, not necessarily to create new content.

Supporting those projects might also require some degree of active experimentation, and are aligned to the mission to benefit readers. The current statement makes it look like the only experiments welcome are those that grow new content or create tools that benefit editors. Diego (talk) 11:51, 17 October 2017 (UTC)

Well, the long form would be more like to build a free encyclopedia that anyone can edit, and the context for zero is to make the encyclopedia more free and allow more people to edit it. In that way, it does align pretty perfectly with the goals of Wikipedia. Besides that, there are, at the end of the day, many many things that are within the scope of the WMF that are not within the scope of Wikipedia. That's why we have sister projects. GMG^talk 12:14, 17 October 2017 (UTC)

I'm not referring to the WMF in particular, that was just an example. I'm talking about a bias from the perspective of editors who are here to build an encyclopedia, and which comes from participating primarily as writers more than readers. For example you above describe Zero as a way to bring more people to edit; but the purpose of Wikipedia is not to edit an encyclopedia, is to have a free encyclopedia that people can use for the knowledge they contain, even people without any will to "build" it. Experiments should be allowed for improving the second use case too.

Guidelines and policies in the long term have an impact on how the community sees the behavior involved. I totally can see how editors in future discussions could use the current wording to support experiments that enhance the editing environment, while rejecting experiments that could largely help how the encyclopedia is consumed by readers, merely because they would be disruptive to the people doing the editing. I want to exorcise that possibility at its roots, by saying something like Editing directed at any goal other than building and using an encyclopedia... thus explicitly allowing the possibility of experiments that improve usage even if they're inconvenient to editors. Diego (talk) 12:27, 17 October 2017 (UTC)

I can see your point. Maybe or improving the experience of editors and readers? GMG^talk 12:36, 17 October 2017 (UTC)

Yes, I like that wording better, I'm adding it to the draft. Diego (talk) 12:38, 17 October 2017 (UTC)

When Should We Post the RFC?

I don't see the need to wait for any other RFC on a different part of What Wikipedia Is Not. Are we ready to post it and to address the reasonable questions? There will be stupid objections, but stupid objections are stupid. Robert McClenon (talk) 02:22, 19 October 2017 (UTC)

Yeah. It's a silly personal preference, but that page may just be too active for it. It's probably not perfect, but it's probably good enough that fixes can be made post facto if it passes. I'm fine with it. GMG^talk 10:06, 19 October 2017 (UTC)

Existing guidelines

There are already some guidelines for researchers interacting with Wikipedia for research purposes, at Wikipedia:Ethically researching Wikipedia#Best practices. I've extended them substantially to address the study that touched all this off, and integrated some ideas from this proposal. Wikipedia:Don't bite the researchers and Wikipedia:What are these researchers doing in my Wikipedia? also contain some relevant content; all three might benefit from a reorg of the subject area.

I've also tried to summarize that study's interactions with Wikipedia. Since the study would have been more successful on its own terms if it had followed community rules and norms better, it seems that the main problem was that the authors didn't understand how to edit Wikipedia.

Almost everything in this proposal is rules that are already well-established (no paid editing, no disruptive editing...), just being ignored by some researchers. So this policy has to solve a communications problem. There are a few possible problems.

For instance, to anyone not familiar with the linguistic conventions of policy shortcut names, it seems as if this policy would ban all research on Wikipedia, rather than just saying that it's not the main purpose and setting limits on it. Could we clarify to "Wikipedia's goal is to build an encyclopedia. Use of Wikipedia as a public laboratory should not interfere with that goal. Editors have volunteered to build an encyclopedia, and not to be experimented upon."?

Secondly, I suspect that the proposal is not an adequate how-to for a new editor. We definitely need some feedback from the target audience before making this a policy, or it could just put the conscientious researchers in fear and uncertainty while not helping at all with careless ones.

Some analyses of Wikipedia's data are not considered OK; wikihounding, for instance. Legally, it's all copyleft, but normatively some analyses are considered unethical by the community.

We already have a way of posting research methodologies at meta:Research. We might want to make this a required form of trial registration, with an alternate proceedure for some trials that might need temporary secrecy. It would be good to have feedback from researchers on this.

It might also be good to clarify what the sharealike expectations are for studies done on Wikipedia. Should the publications be under a open access sharealike license? When would it be required? Wikipedia articles have been published as reviews elsewhere before, so clearly sometimes there is a requirement.

I'm not sure from this proposal how I'd start to go about getting consensus, were I beginning a study. Could we set up some sort of notification system rather than asking a noob to post to one or more of three places? An "I want to do a research study" wizard that asked questions, posted to meta:Research and notified appropriately might be useful.

Comparing studies that caused problems with those that didn't might give some good ideas for guidelines. I'll try and extend User:HLHJ/Draft Wikipedia research experiences to do that, but I don't know the topic well enough to do it alone.

If we rewrote the Wikipedia:Ethically researching Wikipedia guidelines, could we shorten much of the proposal to "If you do interventional research on Wikipedia, follow these [linked] guidelines" and maybe also "to understand some of the things that could go wrong, read [this stuff]"? HLHJ (talk) 06:02, 22 October 2017 (UTC)

Hi HLHJ - my goal with this policy proposal is to have a brief statement that is policy, that is a starting place that lays out the broad principles. Yes policies do interact with each other -- all of them do. This is the place that WMF employees would direct people who want to do research to (researchers often reach out to WMF and there are a couple people there who field those requests) - and yes, this brief, condensed paragraph should contain links to places like the information page WP:Ethically researching Wikipedia that explains things at more leisure, just like RS and MEDRS explains important aspects of V and FRINGE explains important aspects of NPOV. Jytdog (talk) 14:45, 22 October 2017 (UTC)

Most if not all of HLHJ's observations appear correct and salient to me. I think a combined solution would be to work up the guideline well (with material from the essay pages if they seem worth including), then have a very concise NOTLAB policy entry that refers to the guideline for the details. This is typically the sort of thing we do anyway, e.g. WP:AT policy lays out general principles in rather clipped wording, then branches out to all manner of topical naming conventions guidelines for the details, and WP:V and WP:NOR policies say to use (and to not do idiosyncratic synthesis with) reliable sources, while referring to the WP:RS guideline for how to determine source reliability.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  03:12, 23 October 2017 (UTC)