There has been a substantive change to this widely used template [1] based on a discussion held on a different talk page that did not have wide participation. In order to assess if the change is widely supported by the Wikipedia community at large, I am starting a greater discussion here. The RFC question is:

• Does the community support the change in this template from "...should normally follow..." to "...must normally follow..." --Jayron32 12:06, 21 May 2014 (UTC)Reply[reply]

Please input your feelings and rationale below:

Support "should"[edit]

1. The old wording "should" better captures Wikipedia's ethos of no firm rules, and of discussion and consensus taking primacy over rules and bureaucracy. --Jayron32 12:06, 21 May 2014 (UTC)Reply[reply]
2. "Should normally" is better than "must normally" - the latter implies a discrete set of exceptions more strongly than the former. All the best: Rich Farmbrough, 16:21, 21 May 2014 (UTC).
   
3. I waver on this, but we should probably stick with "should". "Must normally" seems to be a little bit of doublespeak to me. If there are exceptions, then why say "must"? The other factor is that our policies are not really written in an airtight, authoritative manner. This is often intentional. If there is a distinction between guideline and policy, it's fuzzy at best. If I repeatedly make bad CSD decisions after being informed of the problem, I could face personal sanctions. But likewise, if I repeatedly brought a bunch of clearly notable articles to AfD, I'd probably face sanctions as well, even though I'd "only" be violating WP:DISRUPT and ignoring WP:N, both guidelines. One could argue that WP:DISRUPT is miscategorized as a guideline, but even if that's so, it highlights the fact that our current documentation doesn't really reflect a good hard distinction between the two. Gigs (talk) 20:21, 21 May 2014 (UTC)Reply[reply]
4. In my view, "should" is sufficiently strong. "Must", even when paired with "normally", conveys a level of strictness beyond that which actually exists.
   In particular, newcomers ideally should adhere to policies in most instances, but they aren't expected to read and understand all of them before editing. They're encouraged to be bold and do their best, with experienced editors correcting their errors and bringing the relevant issues to their attention. The "must normally" wording could be interpreted to mean that any deviation outside exceptional circumstances (even one stemming from unfamiliarity with a policy) is forbidden, resulting in a chilling effect. It's much better that users dive in and learn from their mistakes. —David Levy 17:42, 22 May 2014 (UTC)Reply[reply]
5. Should because WP:IAR is policy and if we write must, then we create circular logic. "You must ignore all rules" and then you must ignore WP:IAR thus you have to follow WP:IAR which then requires you to ignore WP:IAR and so on and so forth. But really because policies need to be fluid and open to interpretation as well as having the ability to be set aside if they are wrong. Policies cannot account for every exception and "should" gives us the room to wiggle.--v/r - TP 19:18, 22 May 2014 (UTC)Reply[reply]
6. Second choice after neither, as WP:IAR means "must" would be deceptive at best. —/Mendaliv/^2¢/_Δ's/ 22:37, 22 May 2014 (UTC)Reply[reply]
7. Support The use of the word “must” implies that policy should absolutely be followed 100% of the time and that a deviation from policy, no matter how justified, is always bad. This implication goes against such established policy as ignore all rules and Wikipedia is not a bureaucracy, as well as common sense and community consensus. I much prefer the use of the word “should.” The word implies that policy should be followed as a default. However, it is less binding than the word “must” and, when paired with the word “normally”, implies that there are in fact situations when policy should not be followed to the letter. This is much more in line with Wikipedia editing standards. Spirit of Eagle (talk) 05:46, 23 May 2014 (UTC)Reply[reply]
8. Per Jayron32, Rich Farmborough, Spirit of Eagle. KillerChihuahua 01:16, 24 May 2014 (UTC)Reply[reply]
9. Should is clearer. The phrase "must normally" is self-contradictory. -- Ypnypn (talk) 17:23, 27 May 2014 (UTC)Reply[reply]
10. Support per Ypnypn. "Must normally" is an oxymoron, since "must" leaves no room for exceptions. We've created a situation in which this template is functionally meaningless, and the only way to fix it (aside from reverting) is to prohibit rule-ignoring and require all policies to be followed to the letter every time. And why would we want to do that? The good of the project is much more important than following policies to the letter all the time. Much better to revert to a meaningful statement. Nyttend (talk) 18:03, 27 May 2014 (UTC)Reply[reply]
11. Support per David Levy and Eagle. "Must" would not be consistent with one of our pillars, WP:IAR. There actually are a few cases (IAR notwithstanding) where "must" may be appropriate (e.g., obey copyright laws) but these situations can use the word "must" on a case by case basis. Rlendog (talk) 15:29, 28 May 2014 (UTC)Reply[reply]
12. Support - I support the pillar and policy of IAR and the word change weakens this. Carrite (talk) 17:39, 1 June 2014 (UTC)Reply[reply]
13. "Must normally" is a contradiction; "must" implies "always". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:10, 1 June 2014 (UTC)Reply[reply]
14. Support just on linguistic grounds. Must normally are juddering against each other: one is high-level imperative; the other works against that. Should is medium-high-level imperative, and idiomatically as well as logically appropriate. Tony (talk) 03:52, 9 June 2014 (UTC)Reply[reply]
15. Support should, based on WP:IAR and that Wikipedia does not have any firm rules. Thanks, Lixxx235^{Got a complaint?} 22:55, 10 June 2014 (UTC)Reply[reply]
16. Weak support on pragmatic grounds. I would mark it "second choice," but for the fact that I don't believe that "must normally," which would be my first choice, has any chance of prevailing. "Normally" alone is insufficient. We can muddle along with the status quo just fine. Regards, TransporterMan (TALK) 12:56, 11 June 2014 (UTC)Reply[reply]
17. Support per WP:IAR, and per the fact that policies are not necessarily thinking through all possible conceivable edge cases. Also "must normally" is self-contradictory. No problems to fix in the current wording.--cyclopia^speak! 02:10, 14 June 2014 (UTC)Reply[reply]
18. Support should - that's the wording used elsewhere, fits to the no-hard-rules and WP:NOTBUREAUCRACY and to apply using reason and common sense. Markbassett (talk) 06:53, 14 June 2014 (UTC)Reply[reply]

Support "must"[edit]

1. Policies carry weight that guidelines do not. Policies include WP:CIVIL, WP:BLP, WP:NPOV, and even WP:IAR. Guidelines include all the other norms (other than essays) that we follow but not as ironclad law. Chris Troutman (talk) 19:15, 21 May 2014 (UTC)Reply[reply]
2. These policies are enforced as a 'must' and this would help greatly. I think making this clear will help reduce disputes on WP by strongly stating that these policies are considered a 'must' with few exceptions. There is always the possibility to change the policy, and there is the piped link to ignore all rules. --LT910001 (talk) 05:08, 22 May 2014 (UTC)Reply[reply]
3. Support. If it's not must, then it's not a policy, but just a guide. Lugnuts ^{Dick Laurent is dead} 12:05, 24 May 2014 (UTC)Reply[reply]
4. "Must normally" is not a contradiction, but leaves room for exceptions in less-than-normal cases.  Sandstein  18:27, 28 May 2014 (UTC)Reply[reply]
5. Second choice. Our policies may not be applicable to a given situation, but compliance is not actually optional. For better or worse, "should" is actually interpreted as "you don't really have to" by a significant minority of editors. IAR does not permit you to reject NPOV. WhatamIdoing (talk) 05:58, 29 May 2014 (UTC)Reply[reply]
6. It's a policy, not a suggestion. And it contains the necessary caveat of "normally" for exceptional cases. --Holdek (talk) 20:54, 31 May 2014 (UTC)Reply[reply]
7. This would certainly cut back on the number of new editors baited into WP:WIKILAWYERING, and as a result, likely improve editor retention... — ((U|Technical 13)) ^{(e • t • c)} 19:41, 8 June 2014 (UTC)Reply[reply]

Support neither[edit]

"This page documents an English Wikipedia policy, a widely accepted standard that all editors must normally follow. Changes made to it should reflect consensus."

1. Fits better "policy is descriptive not prescriptive". If editors don't normally follow, the policy is obviously wrong. The should/must introduces an imperative tone at odds with "community consensus" and "no firm rules". Should/must language begs "or what?". Should/must language belongs with enforceable mediation solutions. --SmokeyJoe (talk) 06:18, 22 May 2014 (UTC)Reply[reply]
2. Well, I like this suggestion and for the reasons SmokeyJoe gives. (BTW the juxtaposition of "must" and "normally" seems very strained to me). Thincat (talk) 16:01, 22 May 2014 (UTC)Reply[reply]
3. First choice. Policies and guidelines are descriptive. SmokeyJoe is absolutely correct. —/Mendaliv/^2¢/_Δ's/ 22:36, 22 May 2014 (UTC)Reply[reply]
4. Second choice. Hadn't thought of this option, but like it. --Jayron32 23:48, 22 May 2014 (UTC)Reply[reply]
5. Distant second choice. While I agree in principle with the rationale, I'm concerned that this might read as too optional, leading to problems and arguments ad infinitim. KillerChihuahua 01:17, 24 May 2014 (UTC)Reply[reply]
6. SmokeyJoe is correct and his proposal fits WP:NOTLAW best. Andrew (talk) 10:26, 24 May 2014 (UTC)Reply[reply]
7. My choice. I echo "or what?". This has the tone of a good faith voluntary obedience. We are volunteer editors after all, but of course, if we misbehave badly enough someone will call us to account. However, we have no General Statutes law book, such as in state or local government, to drag out and thusly handcuff an editor to drag before a magistrate. Fylbecatulous talk 20:09, 26 May 2014 (UTC)Reply[reply]
8. Per SmokeyJoe. -- King of ♥ ♦ ♣ ♠ 05:00, 29 May 2014 (UTC)Reply[reply]
9. First choice. I'd also be satisfied with something like "that editors are expected to follow", as that describes not merely the fact that people do follow them, but that you can expect problems if you don't follow them. WhatamIdoing (talk) 05:53, 29 May 2014 (UTC)Reply[reply]

   I actually like your suggestion a lot. Count that as my first choice. -- King of ♥ ♦ ♣ ♠ 02:34, 31 May 2014 (UTC)Reply[reply]

10. Per SmokeyJoe. Policy should describe common Wikipedia practice and community norms. This is the best wording to suit that. It makes it clear that you should probably follow the rules, but the rules aren't firm and can be changed by consensus. Mz7 (talk) 03:58, 31 May 2014 (UTC)Reply[reply]
11. This one; "must" and "should" are both implying an obligation, with "must" implying that obligation more forcefully; however, it is more an expectation that those aware of policies will follow them rather than an obligation. Bending, breaking, or ignoring a policy in itself is not grounds for a sanction - it is the circumstance, the reason, and/or the continuance that concerns the community. If this option gains consensus, then it would be appropriate to update wording at Wikipedia:Policies and guidelines. SilkTork ^{✔Tea time} 10:43, 31 May 2014 (UTC)Reply[reply]
12. Support because it avoids the problem entirely and I think it is a more accurate description of the way that policies ought to be viewed. (would choose should if this is not accepted, but definitely do not want must) Zell Faze (talk) 19:58, 1 June 2014 (UTC)Reply[reply]
13. I prefer the "editors are expected to follow" version, but this is fine. Neither should nor must really do much for me. There is an expectation that editors will follow policies, however. It makes sense to describe policy in terms of the community's acceptance, and it also makes sense to warn users that they will face opposition when they choose to ignore it. NinjaRobotPirate (talk) 23:22, 4 June 2014 (UTC)Reply[reply]
14. Short, elegant and true. It is not something you are forced to do, it is something we all do. - Nabla (talk) 11:49, 10 June 2014 (UTC)Reply[reply]
15. I like this version too. We could even drop the superfluous and possibly inaccurate "all". All the best: Rich Farmbrough, 03:08, 16 June 2014 (UTC).
    

    Agree with the dropping of "all", as superfluous, inaccurate, tending to hyperbolic rhetoric. These tags should be simple and concise, written to help newcomers, not loaded to support majority domination. --SmokeyJoe (talk) 03:45, 16 June 2014 (UTC)Reply[reply]

Threaded discussion[edit]

• For the record, the previous discussion (linked in the edit summary for the diff above) to implement the change exists at Wikipedia talk:Policies and guidelines#Is a guideline different to a policy?. Mz7 (talk) 20:09, 21 May 2014 (UTC)Reply[reply]

The discussion now exists in the archive Wikipedia talk:Policies and guidelines/Archive 13#Is a guideline different to a policy?. Mz7 (talk) 20:53, 27 May 2014 (UTC)Reply[reply]

• I suggest discarding the "must normally" option at this point. --SmokeyJoe (talk) 00:01, 23 May 2014 (UTC)Reply[reply]
• Lugnuts wrote: "If it's not must, then it's not a policy, but just a guide." This is precisely the sort of misunderstanding that we need to avoid promoting.
  Both guidelines and policies have exceptions. By the same token, neither policies nor guidelines are optional and subject to deliberate defiance on a whim. No matter which label appears, the advice should be followed unless doing so impedes Wikipedia's improvement or maintenance.
  The actual distinction between policies and guidelines is documented at Wikipedia:Policies and guidelines. "Policies explain and describe standards that all users should normally follow, while guidelines are meant to outline best practices for following those standards in specific contexts. Policies and guidelines should always be applied using reason and common sense." —David Levy 21:52, 26 May 2014 (UTC)Reply[reply]
• I'm attracted to the idea that discussions over non-standard behaviour will turn to "Is what this editor is doing good or bad" and opposed to "Is what this editor doing supported by policy". Policy, per se, is not a good reason. Ideally, policy as written, is self-evidently a good idea. Not all policy is written to this standard. Policy, as written, should be informative with regard to practice, and terminology, enabling bystanders to judge whether an editor's behaviour is producing a good outcome. Policy should be subservient to the product, as per Wikipedia:Product, process, policy. --SmokeyJoe (talk) 03:12, 27 May 2014 (UTC)Reply[reply]

  If a policy is written well, applying the advice contained therein will be sensible most (but not necessarily all) of the time. On its own, "because the policy says so" is a poor rationale, but it's helpful to examine why the policy has been written a certain way (i.e. what factors led to such a consensus) and determine whether these reasons apply to the matter at hand.
  In other words, the two questions that you cited aren't mutually exclusive; we should ask ourselves "Is what this editor is doing supported by policy?" and "Is what this editor is doing good or bad?". If the answers don't jibe, we can consider whether this constitutes grounds for revising the policy or for simply making an exception. The mistake is relying solely on whether something is supported by policy. —David Levy 03:44, 27 May 2014 (UTC)Reply[reply]

• Yes. --SmokeyJoe (talk) 03:50, 27 May 2014 (UTC)Reply[reply]
• Agreed. Yet it is often hard to know the "why". I think policy (and guideline) pages would be better server having some kind of "cite your sources" policy. At least pointing to relevant discussions (like this one) so that anyone can get a grasp of the "whys" and the "why nots", and moreover be confident that the current policy text is not just whatever the last editor decided to write down but actually the result of a discussion (and how long, how old, by how many, etc.) - Nabla (talk) 11:55, 10 June 2014 (UTC)Reply[reply]

During the above RfC a suggested wording change to drop the word "should" may not have been clear. The proposal was this: "This page documents an English Wikipedia policy, a widely accepted standard that all editors should normally follow. Changes made to it should reflect consensus" A later suggestion to also drop the word "all" was supported, but came very late. As such there was no clarity on support for that particular suggestion. So I think it's worth considering the proposed wording changes in isolation from the must/should debate. Is there support/opposition for the following wording:

This page documents an English Wikipedia policy, a widely accepted standard that editors normally follow. Changes made to it should reflect consensus

The words "all" and "should" have been removed, leaving the notice wording neutral, informative and factual, with the only obligatory wording being the final sentence, which is the active part of the notice.

This is set up as a simple support or oppose discussion for this template, though it will have implications for the wording on Wikipedia:Policies and guidelines, so this discussion is linked from Wikipedia talk:Policies and guidelines. 05:10, 8 August 2014 (UTC) SilkTork ^{✔Tea time} 13:00, 23 June 2014 (UTC)Reply[reply]

• Support as being more reflective of the intention and practise of policies and guidelines. SilkTork ^{✔Tea time} 13:00, 23 June 2014 (UTC)Reply[reply]
• Support. Reasons stated in the above poll. --SmokeyJoe (talk) 13:20, 23 June 2014 (UTC)Reply[reply]
• Support as more succinct and so clarifying. (Littleolive oil (talk) 16:10, 23 June 2014 (UTC))Reply[reply]
• Support Sounds fine to me. WhatamIdoing (talk) 03:58, 24 June 2014 (UTC)Reply[reply]

Hmmm, a little bit WP:ILIKEIT-y...Forbidden User (talk) 13:03, 8 August 2014 (UTC)Reply[reply]

• Oppose and this should be advertised wider, you don't change policies with discussions on template talk pages without even an RfC. As to the question what happens about 'should' I would point to WP:Policies and guidelines#Enforcement. They are not totally optional. There are consequences and they are spelled out and that is community consensus. Dmcq (talk) 08:56, 25 June 2014 (UTC)Reply[reply]

• I have advertised this as an RfC and at Template:Centralized discussion. Cunard (talk) 05:10, 8 August 2014 (UTC)Reply[reply]

I have listed this discussion as an RfC and at Template:Centralized discussion to increase the level of community input, which is necessary to promote a text to policy. Cunard (talk) 05:10, 8 August 2014 (UTC)Reply[reply]

That's a rather strange reason, given that nobody has any plans to promote this template to policy status. However, I'm happy to have more people join the conversation. WhatamIdoing (talk) 00:13, 17 August 2014 (UTC)Reply[reply]

• Oppose How unclear is the present wording? "Editors normally follow" sounds like a trend describtion more than an instruction. This template is used to infrom/instruct users on the function of templates.Forbidden User (talk) 13:03, 8 August 2014 (UTC)Reply[reply]
• Oppose. I cite the "descriptive, not prescriptive" principle frequently. It means that we put things in writing to document what we do (as opposed to doing things because they're written down), not that we must avoid conveying that anything is required. The tag's purpose is to explain that editors should comply with policies in most situations, not to make a "neutral" statement that this typically occurs (with no indication that it's non-optional). I agree with Forbidden User that the proposed wording comes across more as a trend description than it does as guidance. —David Levy 21:04, 8 August 2014 (UTC)Reply[reply]
• Oppose The use of the word "should" indicates that there is a strong expectation that policy be followed (except in cases where it is appropriate to ignore all rules). If we remove the word "should", then the description makes policy sound like something that editors just sort of happen to do. This does not do justice to the importance of policy, so the word "should" should stay. Spirit of Eagle (talk) 04:13, 10 August 2014 (UTC)Reply[reply]
  • It depends on your perspective. "You should do this–but you don't have to." "People should tell the truth–but they don't." Telling people that all the other editors are doing it, however, is actually a pretty good way to get people to comply. WhatamIdoing (talk) 00:13, 17 August 2014 (UTC)Reply[reply]
• Oppose per Cunard and Spirit of Eagle. The stronger wording should be retained so that it doesn't mislead editors into thinking that policies are completely optional. Otherwise, there is no distinction between policies and guidelines, IMO. Kaldari (talk) 07:30, 11 August 2014 (UTC)Reply[reply]

  I'll note that Template:Guideline also contains the word "should" in this context. Its text simply places greater emphasis on the existence of exceptions. Neither policies nor guidelines are optional (in the usual sense). —David Levy 07:54, 11 August 2014 (UTC)Reply[reply]

• Oppose  Removing the word "all" is ok, but removing "should" weakens the language.  Unscintillating (talk) 01:00, 12 August 2014 (UTC)Reply[reply]

  I agree that the word "all" can be removed. It makes the sentence shorter and more concise without losing meaning in the process. Spirit of Eagle (talk) 02:16, 12 August 2014 (UTC)Reply[reply]

• Oppose - the word "should" is correct - it's weaker than "must" (an absolute requirement), which is clearly contradicted by IAR, but string enough to mean that users are expected to do it for the most part. עוד מישהו Od Mishehu 08:12, 13 August 2014 (UTC)Reply[reply]