Misplaced Pages

This is an old revision of this page, as edited by Neotarf (talk | contribs) at 12:24, 19 September 2012 (→Should: format). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

This is the talk page for discussing improvements to the Policies and guidelines page.

Put new text under old text. Click here to start a new topic. New to Misplaced Pages? Welcome! Learn to edit; get help. Assume good faith Be polite and avoid personal attacks Be welcoming to newcomers Seek dispute resolution if needed

Archives: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Auto-archiving period: 7 days

The project page associated with this talk page is an official policy on Misplaced Pages. Policies have wide acceptance among editors and are considered a standard for all users to follow. Please review policy editing recommendations before making any substantive change to this page. Always remember to keep cool when editing, and don't panic.

This is the talk page for discussing improvements to the Policies and guidelines page.

Put new text under old text. Click here to start a new topic. New to Misplaced Pages? Welcome! Learn to edit; get help. Assume good faith Be polite and avoid personal attacks Be welcoming to newcomers Seek dispute resolution if needed

Archives: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Auto-archiving period: 7 days

Misplaced Pages:Religion

The page Misplaced Pages:Religion has been marked as failed... however there is an ongoing tlak about the tag I added today, thus was reverted. I added the tag to indicate the page is being surpassed by the new proposal at Misplaced Pages:WikiProject Religion/Manual of style. Can we get a few more people to comment on the old page pls ... see Misplaced Pages talk:Religion#Failed proposal from 2009.Moxy (talk) 17:30, 4 August 2012 (UTC)

Should

What do you all think about adding an external link to http://tools.ietf.org/html/rfc2119 (the famous RFC 2119 that defines the differences between words like "should" and "must") to this page? WhatamIdoing (talk) 21:08, 4 September 2012 (UTC)

It's already linked in Misplaced Pages:Policies and guidelines#Content - possibly you meant something else? Or repeating it at the bottom? Or linking to it with additional context, such as "We have NOT gone through all policies and guidelines to make sure they conform to a legal-eagle level of defensibility, especially with regards to word choices such as these, which well-intentioned editors are often changing on any given day." ;) -- Quiddity (talk) 00:28, 5 September 2012 (UTC)

I think that we need more than just a link. I think that we need to have something that explains the terms (briefly) and recommends that using them precisely is the best practice. Perhaps something like this: "Use the word must to indicate that something is mandatory. Use the word should to indicate that something is appropriate and recommended under nearly all circumstances, but not actually required." WhatamIdoing (talk) 02:46, 12 September 2012 (UTC)

The trouble that the document referred to has no authority for Misplaced Pages use. "Should" and "must" are not distinguished as that document stipulates: neither in actual practice on Misplaced Pages, nor in natural usage generally. This can be confirmed by checking standard dictionaries (including OED) and the major descriptive grammars published in recent decades (such as CGEL, the pre-eminent current grammar of English). WhatamIdoing, weren't you the one who inserted the Easter egg link to an external document into policy, without explaining in your edit summary what you were doing? I think it should be resisted. If you want such an appeal to an external "authority" that runs counter to the current usage of all policy and guideline pages on Misplaced Pages (and also the usage we see in ArbCom decisions, for example), please present the proposal in a neutral RFC, so that the community can assess the value of such a radical innovation.

☺

Noetica 03:59, 12 September 2012 (UTC)

I don't mind setting up an RFC. The purpose of that section is to provide "best practices", rather than an accurate description of the overall sorry state of advice writing to date. Given the responses at WT:AT, I expect that the community will support defining these terms this way, but I have absolutely no objection to an RFC on the matter. What do you think about this style:

No, we should not make this change	Yes, we should make this change
We should not make this change because: We've used should to mean must for years, so we should stick with our old practices. The word must scares people.	We should make this change because: It confuses people when should sometimes means "if you feel like it" and other times means "with no exceptions". Editors need to know what's mandatory (like removing copyvios) and what's not (like notifying editors of AFDs).

What do you think of that style? Would you like to write the "anti" statement for it? WhatamIdoing (talk) 19:48, 15 September 2012 (UTC)

That looks good/reasonable to me (if the "anti" statement can be worked on and agreed to. Ie. get the wording to a consensus-neutral before posting).

Note to anyone not watching the other thread, the WT:AT responses about this are at the very bottom of WT:AT#Discussion (Search for "Presenting"). —Quiddity (talk) 20:13, 15 September 2012 (UTC)

WhatamIdoing, thank you for the work you have done on this. Some responses to comments above:

• Your proposed style of presentation is fine, except that we lack details of the exact proposal. I'd need to see that before making a definitive judgement.
• The provisional text you have in the presentation is not yet satisfactory, of course. One problem: you use the word "should" even in presenting the options! Strangely, this might be distracting and confusing. It is essential to keep all the issues lucid and separate. Also, the text is based on presuppositions that compromise neutrality and clarity. This, for example:

"It confuses people when should sometimes means 'if you feel like it' and other times means 'with no exceptions'."

This presupposes that the variation mentioned does actually occur, in quasi-legislative contexts where policy or guidelines are presented. This is not so, except in expressions such as "if an editor should revert". These are quite distinct from the core deontic uses of "should", and a complete red herring.

• These matters are technical, and should be presented in a way informed by sound linguistic theory. This has not been the case at the WT:AT RFCs. Those have been disastrously confused. It is not at all certain that the community would want to change how we use "should", from discussion at WT:AT.
• I would be happy to draft a "no" case: that is, the case against distinguishing "should" and "must" according to the stipulation of that isolated external source. I would also be happy to assist with technical advice, to achieve the essential clarity and precision. Please call in at my talkpage, if you would like that. ☺

Noetica 00:21, 18 September 2012 (UTC)

I think the next step is to create the question, since it will be difficult to write a sensible case for or against it if we don't know the proposition. Do you like the proposed text (in green serif above), or is there another change that you would like to discuss? WhatamIdoing (talk) 00:36, 18 September 2012 (UTC)

I like the use of green xt markup, but not that wording. Frankly, I don't know how this can be managed effectively: if legacy uses of "should" would mostly be left as they are, and the new policy provision would be picked up mainly for new text!
Let us see precisely what you have in mind: accounting for the weight of that legacy, and for inevitable and often undetectable non-compliance with the new provision you propose. The core problem: How would guidelines and policy be interpreted, if it is not known what is legacy and what is new, and not known what complied with and what ignored the new distinction you propose? Compare the recent and continuing chaos over "should" in that infamous WP:AT provision (subject of two inclusive RFCs so far). Much easier and safer to treat both "must" and "should" as having equally clear deontic intent, whatever their differences in tone.
We're tinkering in Misplaced Pages's CNS here.
Noetica 01:17, 18 September 2012 (UTC)

I object to this entire process. First of all, Misplaced Pages has its own internal rules and standards, policies and guidelines. If you say we're going to use some obscure Document X as an authority to override policy, in contravention of standard English usage, then what's to stop anyone from elevating Document Z, or the CMOS, or the amusing and much-cited Elements of Style as an authority over MoS? You think there are quarrels now over style guides, just wait until everyone brings in their favorite pamphlet or memo from work and wants to apply it across Misplaced Pages as a whole. Second of all, no one has shown that there is any true confusion over "should". It is not all that uncommon that people's description of their usage is different from their actual usage; that's why linguists use tape recorders instead of self reporting. Even the people who have claimed to be confused have had no problem using and interpreting the word correctly. 3) "Must" is British; Americans prefer "have to". Both of these are the proper register for talking to naughty 5-year-olds, this is not the proper register for an encyclopedia or for corporate policy. 4) The degree to which anything in Misplaced Pages is mandated does not rely on specious definitions. The role of policies and guidelines is already defined. Neotarf (talk) 19:54, 18 September 2012 (UTC)

1. Not a geek, are you? Far from being some obscure, unimportant outside document, RFC 2119 is one of the reasons the Internet works. It is so basic that it's a WP:Magic word on WMF websites. It also does not contravene standard English usage. (Did you read it?) Its definition of should fits neatly within Merriam-Webster's fourth, "used in auxiliary function to express what is probable or expected". This is not the only possible definition of this word, but it's the only relevant one for rule-making. Its definition of must is similarly compatible with the dictionaries I've consulted. And, importantly, none of the dictionaries I've seen so far believe that must is a valid definition of should, which is the particular dispute that we're dealing with.
2. Have you forgotten the recent mess at WP:AT already? Noetica just spent a couple of weeks trying to tell everyone there that when that policy says "Any potentially controversial proposal to change a title should be advertised at Misplaced Pages:Requested moves" it actually means "Any potentially controversial proposal to change a title must be advertised at Misplaced Pages:Requested moves". We definitely have confusion caused by undefined use of these terms. If we mean that RM is required, we need to say that, and not hide that requirement. And if we don't mean that it's required, we should put editors on notice that when we said should, we actually meant should instead of must.
3. This particular American prefers the term must over the term have to. It's possible that I'm not the only one, either.
   The proper register for corporate policies is the one that clearly communicates the desired behavior, in ways that don't leave reasonable people drawing opposite conclusions. The proper tone is whichever tone doesn't result in Noetica and a handful of other people having a 200-kb-long tis-tisn't "discussion" over the meaning of the word should in a particular sentence. The proper register for corporate policies is also one that doesn't result in Noetica (or anyone else, but that's our current example) opening ANI and ArbCom requests over a disagreement about whether something that "should" be done absolutely "must" be done with no exceptions.
4. We're not talking about specious definitions. We're talking about expanding the existing text about using these terms to encourage everyone to use them with precision and consistency, so that when a policy says "Any potentially controversial proposal to change a title should be advertised at Misplaced Pages:Requested moves", nobody will wonder whether that means "absolutely required, and you are entitled to start endless discussions and request ArbCom sanction everyone else if it's not done" or if it means "it's a good standard practice, but there are some legitimate exceptions". WhatamIdoing (talk) 21:46, 18 September 2012 (UTC)

Disappointing, W. Shameless personalising of a key issue in the interpretation of policy and guidelines throughout the Project. Don't believe the loud and indignant rhetoric. I was not the one! I did not go to WP:AT while I was involved up to the hips, unilaterally overturning a plain provision that had been serenely and explicitly in policy for three years, with a wikilawyering quibble over the meaning of "should". I did not then start a stupidly disruptive RFC that had no structure and no hope of resolution, wasting days of editors' time and peppered with confusions so basic they would be dispelled in the first week of Linguistics 101. For the rest – your risibly misunderstood evidence from M-W, your appeal to a geeky backroom text that rests on equally flawed assumptions about real English – it is not worth going through all that now. If you want an RFC, frame it well and fairly – or I will well and fairly object. Someone has to! If despite appearances you are ready for expert assistance rather than piling on the populist calumnies, just let me know, OK? ♥
Noetica 22:13, 18 September 2012 (UTC)

Noetica, I am aware that there are some legacy problems with our policies. It's taken a couple of years just to clean up the mess caused by people who didn't understand that WP:Secondary is not another way to spell good. But if we decide to define these terms, then we can clean up the policies, and as a bonus, you'll (eventually) never again have to wonder whether any given instance of "should" actually means "must".

My question about the green text is whether that's the question you want the RFC to focus on. We need a brief question. One option is

"Misplaced Pages:Policies and guidelines#Content's existing advice on writing policy and guideline pages says "be clear. Avoid esoteric or quasi-legal terms and dumbed-down language. Be plain, direct, unambiguous, and specific. Avoid platitudes and generalities. Do not be afraid to tell editors directly that they must or should do something." Shall we add the following text to that item?

"Use the word must to indicate that something is mandatory. Use the word should to indicate that something is appropriate and recommended under nearly all circumstances, but not actually required."

Does that sound like a fair and brief question to you? WhatamIdoing (talk) 21:52, 18 September 2012 (UTC)

(The sequence of these posts got mixed up. A server problem? Somehow I did not see that post before I posted my own last one. Anyway, it stands.) No, W. That's not going to work. You should first remove the aberrant link from core Misplaced Pages policy to an external source, which on a charitable interpretation you neglected to signal in an edit summary when you inserted it. Then the whole issue could be looked at freshly. Your proposed wording for the RFC question does nothing to address the questions of implementation, compliance, and interpretation that I outlined earlier. You may think we can "clean up the policies" on the basis of some 1997 reworking of the English language; I do not share your optimism.
Noetica 22:27, 18 September 2012 (UTC)

Doubleplusungood. What's next, a specially defined alphabet? Hmm, here's Merriam-Webster. Definition 2: "—used in auxiliary function to express obligation, propriety, or expediency". Neotarf (talk) 23:50, 18 September 2012 (UTC)

Neotarf, apart from several descriptive grammars and other truly relevant resources, I have surveyed a dozen dictionaries on this topic. The M-W treatment of "should" is the worst I have seen, mostly because the citations are very poorly sorted. From the Collegiate, which you and WhatamIdoing appear to be citing:

5 used in auxiliary function to express a request in a polite manner or to soften direct statement {I should suggest that a guide ... is the first essential – L. D. Reddick}

Does Reddick's "should" express a request? No, not by itself. If anything, the verb "suggest" does that work, and "should" (functioning very like entirely non-deontic "would") adds to the circumlocution. Does that "should" soften direct statement? Possibly. We would need to see the context. It might turn out to be conditional: " I should suggest ...". The fuller version under this heading from Webster's Third New International (big sibling of the Collegiate) adds these citations (and the lemma is now numbered 6, not 5):

{one aspect of his critical work to which I should like to call attention – Malcolm Cowley}  {should you wish to look at it – O. Henry} {in general I should say that the salaries ... make up very nearly two thirds of the budget – Deems Taylor}

In these "should" plainly does not itself "express a request". Not unproblematically. Does it "soften direct statement"? Yes, but only in a way that is remote from the expression of obligation. Certainly not in any way that is germane to contexts such as this statement from WP:AT: "Any potentially controversial proposal to change a title should be advertised at Misplaced Pages:Requested moves, and consensus reached before any change is made." And that O. Henry citation is squarely conditional, equivalent to "if you wish to look at it" or the awkward "if you were to wish to look at it" (for a conditional with irrealis mood). It is therefore misclassified under that lemma. It belongs under 1, which in the Collegiate version is this:

1 —used in auxiliary function to express condition {if he should leave his father, his father would die – Genesis 44:22 (Revised Standard Version)}

When we add people's misreadings of M-W, we glimpse the scale of the problem. Those are all distractions from quite separate uses of "should", where obligation is clearly to the fore.

These things are better treated in a properly conducted RFC – if we can still hope for such a thing, after recent examples of that genre in decline. ☺ I don't want to waste more time on them.

Noetica 01:07, 19 September 2012 (UTC)

Noetica's deep immersion in his learned sources reminds me of a person who drives into a ditch because according to the map, there is no ditch there. IMO, it is incontrovertible that should is ambiguous for a significant portion of the population. If the goal is to indicate an obligation for which one will be subject to sanctions for ignoring, then "should" is not the right word to use. older ≠ wiser 01:36, 19 September 2012 (UTC)

Indeed... in fact, I would go a step further - in most cases where we do use the word "should" in policy, we chose that word because it is a somewhat ambiguous word. In most cases, what we are describing in the policy is "Best Practice" not a mandated rule. We are intentionally not saying "must" because we recognize that the normal "Best Practice" being described has exceptions... situations where it makes sense to do something else instead. So, in those few situations where we really do mean "must" we should use that word instead. (or, better yet, use the active voice and simply say "Do X"... or "Don't do Y"). Blueboar (talk) 02:31, 19 September 2012 (UTC)

≠ (alias User:Bkonrad):

• Better to be "immersed in learned sources" (as I cheerfully admit that I often am) than to be unable to parse a sentence when challenged to support one's reading of it (as you were when we last discussed all this, in a doomed RFC).
• Yes, the plainest deontic reading of "should" (usual in policy and guidelines, as in any quasi-legislative context) can be distorted, and can be misunderstood. Sure! But the solution is not to assume that it is has a less common and typically stipulative meaning. The solution is to survey the issues without prejudice, and adopt an approach informed by sound analysis of the language as most people use and understand it.

Blueboar:

• You make an interesting claim, without support, about why "should" is typically chosen. In a proper RFC, I will show another reason for people preferring to use "should", still with full obligatory force. Backed by empirical evidence from a corpus-based grammar of English.
• I agree with you about using direct instructions instead. It solves a multitude of problems in policy and guidelines. But you misidentify that as active voice. It is no such thing. Most "should" and "must" constructions are in active voice! (Not the one that occasioned all this wrangling, though: "Any potentially controversial proposal to change a title should be advertised ...".) I think you mean imperative mood. Not a bad idea to check those "learned sources" from time to time, right? ♫♪

Noetica 03:18, 19 September 2012 (UTC)

Noetica, (and I have lost track of the threading here), not the Collegiate, the online version of M-W is identified as the "Eleventh Edition". It has been superseded of course, that's why it's free, but there is much readily accessible information there, -- did I mention that it was free? Many of the examples you cited (and also from the online version) seem like they are no longer in common usage, at least not in my neck of the woods. The second definition certainly applies to policy:

—used in auxiliary function to express obligation, propriety, or expediency <'tis commanded I should do so — Shakespeare> <this is as it should be — H. L. Savage> <you should brush your teeth after each meal>

Clearly "should" can be used to express obligation, such as the obligation to follow policy. "Policies have wide acceptance among editors and describe standards that all users should normally follow." "Normally" here links to WP:IAR; clearly "must" would be inappropriate here. Does the markup code cease to work if policy is not followed? No. One can choose whether or not to fulfill obligations, but of course, either way there are consequences to be calculated.

Can anyone mistake policy for being a mere recommendation just because it uses the word "should"? I think not. The guidelines are recommendations. "Guidelines are sets of best practices that are supported by consensus. Editors should attempt to follow guidelines, though they are best treated with common sense, and occasional exceptions may apply." The difference between "should normally follow" and "should attempt to follow" are much clearer descriptions than arbitrarily redefining words in some obscure back page somewhere. Neotarf (talk) 12:07, 19 September 2012 (UTC)

W: I support your proposed addition to #Content but I'd prefer the phrase "not mandatory" or "not obligatory" instead of "not actually required" because it more clearly contrasts "should" usage with "must" usage and "actually" makes me wince a little, actually. I think this is a useful clarification of long standing usage/interpretation of "should" and "must". Has anyone put together an analysis of "should" usage on policy pages? Jojalozzo 02:43, 19 September 2012 (UTC)

Y'all can find out "how we currently use 'must' and 'should'" most efficiently, by downloading either Book:Key Misplaced Pages Policies & Guidelines or Book:Be Misplaced Pages Wise (and/or making/updating a book). There are 402 uses of "must", 2105 uses of "should", in that second book/collection. HTH. —Quiddity (talk) 05:56, 19 September 2012 (UTC)

Misplaced Pages namespace redirect to essays in user space

I have noticed some essays have shortcuts in Misplaced Pages namespace that redirect to essays in user space. This feels odd to me for several reasons. It seems to imply some degree of endorsement of an essay as "official" yet still maintaining some degree of "ownership" of the essay by the original author. It also prevents competing essays on the same topic from using what may be the only logical shortcut. Is this something that editors should not do? Jason Quinn (talk) 07:27, 10 September 2012 (UTC)

There was a recent proposal to link to essays through shortcuts with the E: prefix. You may want to relaunch it. Diego (talk) 08:30, 10 September 2012 (UTC)