Wikipedia talk:Manual of Style/Linking/Archive 21

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 15 ← Archive 19 Archive 20 Archive 21

Proposed change: Allow linking once per section, not just in first section after lede?

The current policy states: "Generally, a link should appear only once in an article, but if helpful for readers, a link may be repeated in infoboxes, tables, image captions, footnotes, hatnotes, and at the first occurrence after the lead. Citations stand alone in their usage, so there is no problem with repeating the same link in many citations within an article."

However, on articles this is often a problem for me, especially long articles, because I might be reading just a section of the article. On occasion I have to actually go searching in the rest of the article to find the wikilink. This is annoying and a pointless waste of my time! One time I even added a wikilink because I looked for it and couldn't find it, and it was removed because it was elsewhere in the article - just in a section I hadn't read (because I was only interested in that particular section!)

Wikilinks are cheap - why not relax this and allow them once per section?

Proposed re-wording:

"Generally, a link should appear only once in an article, but if helpful for readers, a link may be repeated in infoboxes, tables, image captions, footnotes, hatnotes, and at the first occurrence in a section. Citations stand alone in their usage, so there is no problem with repeating the same link in many citations within an article."

This might cause over-linking but given the "if it is useful to readers" language, this allows it when it is useful, and overlinking where it is not necessary (i.e. if there are very few sections or if the sections are very short) could still be removed.

Mvolz (talk) 08:54, 3 May 2021 (UTC)

Oppose because "section" is not clearly defined. I have seen articles where a level-two heading is broken down into multiple "sections" by level-three headings and each of these sections is one sentence long. I have also seen articles where there are multiple one-sentence paragraphs broken up by level-two headings.
At best we could say that there may be a link every 300 words, provided that there is at least one new level-two heading between the two parts of the article. Walter Görlitz (talk) 18:46, 3 May 2021 (UTC)
That could definitely work but it seems a bit overly complicated. What about specifying once every top level section? (i.e. heading 1?) Mvolz (talk) 09:22, 4 May 2021 (UTC)
It really needs to be per screen page. As I wrote, I have seen articles where there were one sentence sections. Walter Görlitz (talk) 05:11, 5 May 2021 (UTC)
I agree per screen page is in fact of the matter ideal. Could you word that into a linkin policy though? I imagine if we actually said that we'd end up in a screen size quibble because there's mobile all the way up to huge LCD monitors! I have also seen articles with one sentence sections, but if it's a *top* level section it's usually because it's a stub... and if it is overlinked, you could still remove the links because the phrasing of the policy says "if it is helpful for readers." The policy still says only once in an article is the main policy. Mvolz (talk) 06:43, 5 May 2021 (UTC)
That's the point. A screen page differs depending on the size of screen and browser on it. I have multiple monitors here, and so it's difficult to quantify. When I worked for a web company, we had a standard minimum screen size for which we designed, but I have never seen the same for Wikipedia. Also, it would change over time. When Wikipedia started, 800x600 would have been considered a normal monitor and within a decade, 1024x768 supplanted that. Now we can see much larger. That's why I was suggesting a "words of rendered prose" metric. By the way, we are talking about REPEATLINK, not OVERLINK. Walter Görlitz (talk) 06:48, 5 May 2021 (UTC)
MOS:REPEATLINK is a guideline, not a policy. It's a suggestion. You're free to ignore it. I do, and I've never been gainsaid, and I've basically never come across it presenting a problem. I suppose someone might come along and, citing the guideline, remove them, altho I've never seen that. If someone does, I suppose it keeps them from making mischief somewhere else, anyway.
I do it when the term pops up again after we've had lengthy material on something else, or in first use in a section that readers are likely to be jumping to directly from the article menu. Which is many, but not all, sections, falling back on common sense and best guess.
Right, as Walter Görlitz said above MOS:OVERLINK -- linking common terms and so on -- is entirely different, and is a common problem. Herostratus (talk) 07:14, 5 May 2021 (UTC)
@Herostratus, in practice, when editors take long articles to FAC, and the prose links to the same article once in the lead and twice in the body, reviewers at FAC have insisted that the last link be removed, and that the first link after the lead needs to be the first occurrence of the word after the lead, because otherwise it's a MOS violation. If you don't think that's appropriate, then this guideline needs to use different words – words that cannot be construed as meaning that a link can only be repeated once per article (no matter how long) and that it must be the first occurrence of the term after the table of contents (even if another section would be more pointful). WhatamIdoing (talk) 16:21, 18 June 2021 (UTC)
[[User:WhatamIdoing, I don't know anything about FAC. People there do what they think best I am sure and that's fine and God bless them for their work, but I can't control what they do either way.
If serving the reader means linking more than once, I'll do it -- and recommend that you do too. There are millions of articles here, so fair chance nobody will find it and mess with it. If they do, oh well, you tried. If the article gets sent to FA and thus people there remove the links, oh well, it is what it is. In the meantime you served all the readers who got to the article before that, at least. (I'm sure FA is mostly great and mostly improves articles, but I get that to make it work without constant discussion of every point you want to have some written standards to agree on, and fine, but the law is a blunt instrument and I'd be surprised if there wasn't sometimes a Procrustean effect i play.)
"then this guideline needs to use different words" may be true, but I can't make it do. Maybe somebody else can and godspeed to them. It's not so easy, since as you can see even in this thread it's a common human trait to want to boss other people around, to be unable to understand the experiences and needs of people different from oneself, to feel comfortable in an environment of written-out rules, and like that. Immediately below me I see an editor writing "Above all of this is some nervous notion that readers can't tap a target into their search box in eight seconds". You can't work with people like that. Best is to just ignore them and hope they don't notice you. This lubricates the project as it does most organized human undertakings. Herostratus (talk) 17:20, 18 June 2021 (UTC)
The appeal to feature article is specious. Feature article reviewers simply try to follow guidelines and manuals of style they know about. I have seen atrocious looking feature articles that ignore guidelines but have great spelling and grammar, and others that are the reverse. Walter Görlitz (talk) 06:02, 20 June 2021 (UTC)
I don't think we're as helpless as all that. The FAC folks are trying to comply with the MOS. When complying with the MOS causes problems in articles, then we change the MOS to make a better recommendation. They'll then do their best to comply with the better rules. WhatamIdoing (talk) 16:01, 20 June 2021 (UTC)
  • This would be a terrible move. Why stop at once per section. What about once per paragraph? Once per sentence? Above all of this is some nervous notion that readers can't tap a target into their search box in eight seconds. NO. Tony (talk) 04:09, 6 May 2021 (UTC)
    • Why don't you actually make an argument as to why it is better to make navigation more difficult for reassess instead of engaging in a slippery slope fallacy? oknazevad (talk) 11:51, 21 June 2021 (UTC)
  • "On occasion I have to actually go searching in the rest of the article to find the wikilink" It might seem elementary, Dear Watson, but if ever I feel the need to do that, I wouldn't blame the lack of links on someone else "wasting your time". I would simply input the term directly into the search box. You might be surprised how well that works -- Ohc ¡digame! 08:28, 6 May 2021 (UTC)
    I am well aware of the search page, but on many occasions wikilinks in the article are more substantially more useful. In the one particular case I'm thinking of, it was a person's name, of which multiple such people with the same name exists. With wikilinks the disambiguation is already done for you. Or, in any case where the subject doesn't match the exact title of an article (this is common in medical articles where it's often necessary to translate jargon into lay terms or vice versa). To argue it's no less easy to find the article via search than to search the page directly or to directly click the wikilink is an argument against wikilinks altogether. Mvolz (talk) 07:20, 15 June 2021 (UTC)
  • I am often annoyed when things are not linked in the section I am reading (which may be the first bit of the article I see thanks to redirects to sections). It should not be discouraged to repeat highly relevant links to nontrivial targets. (Of course an article about a French politician should not link to France more than once, if at all). —Kusma (t·c) 08:33, 6 May 2021 (UTC)
    • If you parachute into the middle of a text, don't expect not to have to scroll up to get the context. Every extra link waters down the value of the linking system (which is not used nearly as much as some editors think it is); and makes the text blotchy and slightly harder to read smoothly. Ration, please, has been the system for significantly more than a decade. Tony (talk) 09:32, 6 May 2021 (UTC)
      I don't volunteer to parachute into the middle of a text, it is something that redirects to section do to me, and then i assume I'm supposed to read from where I've been redirected to instead of from the top. (I don't like redirects to sections much). I am all for rationing the number of links, but still want to caution against a hard "no page may be linked to twice". Another case for second links is when the first link is hidden behind a pipe that isn't totally obvious. —Kusma (t·c) 10:11, 6 May 2021 (UTC)
  • If anything on this is to be added, it shouldn't endorse repeating links on the same level as for infoboxes, tables, captions, etc. I don't believe we need to be catering for readers who drop into a section midway through an article. (That goes for me too: if I jump from the ToC to a particular section of interest, I am going to find surnames and wonder who we're talking about, acronyms that aren't defined, and terms that I might wish are linked – but I've made the choice to engage with the text in that way.) Having said that, it's hard to avoid having repeated links in long biographical articles, especially for very well-known or successful individuals; I don't add them, but I see some sense in it when others do. I'm thinking about when, late in the article, there's a section on the subject's personal life, and partners and children are named. For actors, that might include co-stars that have been introduced in sections discussing their career. In articles about music and musicians from the 1960s and '70s (which is the area I almost exclusively work on), a wife or girlfriend has invariably been discussed already, perhaps as inspiration for the artist's music, sharing or inspiring the artist's political or spiritual preoccupations, getting arrested together ... JG66 (talk) 08:47, 6 May 2021 (UTC)
  • I call these sort of issues “pendulum swings from one extreme to the next.” They occur in business, occupational safety, real life, and Wikipedia. It can be someone who presses the limits in some way, or performs a task in an unsafe manner, or is way off in left field with… anything, and there’s an unfortunate outcome followed by a knee-jerk reaction.
Such incautious actions can be over-linking the living daylights out of a Wikipedia article… and repeatedly linking the same thing over and over. Or it can be someone from the Maintenance department falling off the third step of a ladder. In the case of over-linking, the pendulum swing can be “use a link once and only once in an article” and in the case of ladder safety it will be “all 6000 employees in a company (not just workers from Maintenance who change light bulbs) must have an hour of Ladder Safety Training quarterly.”
I don’t think there is a single, hard, fast rule that can properly apply to repeating a link in an article. The challenge is wikipedians can be everyone from a 7th-grader taking a first crack at editing an article to a professional author with 15 years of experience on Wikipedia. I, for one, try to link once per article. But, depending upon circumstances like these:
  1. the uniqueness of the term for the expected target audience,
  2. the way the article is organized and segmented (it may be highly and necessarily linear or it may have semi-associated subtopics that will often be skipped),
  3. the subject matter (it might be exceedingly technical),
  4. the target audience,
…I will sometimes repeat important links.
The Wikipedia project is too vast and the subject of linking is too complex to have hard-fast rules. MOS, in my opinion, should lay down general guidance on key aspects and principles of linking and why one does so, explain the underlying rationale for why the guidance is as it is, and provide some example uses. Being overly prescriptive one way or another in anything in this world seldom works well. Knee-jerk reactions are easy and fast to implement. However, creating nuanced, thoughtful, flexible guidance that works in the longterm is double-tough and requires effort. Greg L (talk) 14:59, 6 May 2021 (UTC)
You, with your ... reasonableness and common sense. People like you make me sick! EEng 16:42, 6 May 2021 (UTC)
:-) (Greg L (talk) 00:50, 7 May 2021 (UTC))

twice per body if long - I think we could simply relax it a bit. We can generally recognize overlinking -especially in egregious sea of links articles-, but it is not completely simple to recognize the point where it would be useful to add it again. Myself, I would allow a second in the body if the article is long, but not two wlinks in the same top-level section. When I remove extra links, I find that often that second links or more are close together. I attribute this to different editors adding a new sentence and not noticing the previous wlink. Or rewrites. I would not add too much complexity to the rule. Let's not swing too far from the one per body rule. Alaney2k (talk) 13:36, 8 June 2021 (UTC)

I support this proposal by Mvolz for one link per ==Section==, for whichever words editors feel would be helpful to people who are reading that section. The only slight difference I have with the proposed wording change is that I wouldn't bother specifying that it must be specifically the "first occurrence". I agree that it is not at all unusual for people to skip to the section that they are most interested in, and thus miss key words. (Imagine that you're reading an article about chemistry, or some other technical subject. The existence of a link in a section you're not reading is not helpful at all.) Particularly with the ability to hover over a link and get a quick definition, I think that greater flexibility and (within the bounds of what editors agree to on the individual article's talk page) is a good idea. On a related point, this problem was discussed at Wikipedia:WikiProject Medicine recently. I didn't know this discussion was underway, or I would have linked to it. I would particularly like to see us stop stripping links to technical terms out of medical FAs just because "the rules say so". WhatamIdoing (talk) 05:03, 15 June 2021 (UTC)
That will soon lead to a sea of blue, disrupting the reader and cloggin the appearance of the text. Not only that, it will dilute the hyperlinking system. Our responsibility as editors is to ensure that linking is rationed to the most useful. If one in 10,000 readers happens to want to divert to another article in a particular section, they can so easily type the name into the box. It's that easy. Tony (talk) 05:53, 15 June 2021 (UTC)
It's that easy... for those of our readers who are on a desktop computer. Try to use the mobile view: by default it shows sections as collapsed. The MOS must cater to all of our readers (and a huge number of them are on mobile), and making our articles more usable may be worth dilut[ing] the hyperlinking system. Overlinking is bad, but underlinking is worse. —Kusma (talk) 08:59, 15 June 2021 (UTC)
It's that easy… if you're done with the article you're currently reading, or you don't mind losing your place on the page. But if you just want a quick reminder about which the article is talking about the Anacin with caffeine or the Anacin-3 without it, or the hydr-something-zine that's for blood pressure or the similarly named drug for allergies, then stop reading the article so you can go separately look up one word is suboptimal. WhatamIdoing (talk) 21:31, 15 June 2021 (UTC)
I think we do have to find some balance between the over-restrictive "once per article body" and the "sea of blue" that Tony fears. Btw, the guideline text is currently just plain wrong for articles with a comprehensive lead that is a summary of the body: the lead and the body are permitted a link each, and yet we continue to introduce this as "once per article". I disagree that "It's that easy" is in any way an acceptable argument. If there was no loss in removing hyperlinks, because one can always copy/paste or type into the search box, then we could just remove all hyperlinks, and rely on the search box. So clearly a hyperlink serves a useful purpose. Not only for convenience but also to make a statement that Wikipedia has an article on that, which perhaps you'd be interested in reading? A large part of the enjoyment of reading Wikipepdia articles is the rabbit hole one ends up down as one follows links and learns about things one never intended to read about.
The "once" rule makes a huge assumption about reading patterns and about how much attention readers spend reading each section. If one is skim-reading or simply skipping sections entirely, then the earlier hyperlink is either not remembered or not even visible. So it might as well not have been there. That is the problem we should consider a solution for. I think this is most important for complex terms or words or events or things where the reader is most likely to need help, or to investigate further. When a reader comes across a complex term that is not hyperlinked, they may receive a really bad message: the writers of this text though you already knew all about this word, so didn't feel the need to help you with it at all... clearly you are too dumb to be reading this, so I would give up now if I were you. -- Colin°Talk 09:42, 15 June 2021 (UTC)
I agree that we need to reach a middle ground. I don't think we will reach a "sea of blue", and I don't think anyone actually believes that would happen. It's just the usual hyperbole, intended to show a concept rather than a realistic outcome. WhatamIdoing (talk) 21:35, 15 June 2021 (UTC)
Strong support. Current MOS statements are actively counterproductive to writing articles readers can expect to understand on any topic with jargon. "It'll devolve into overlinking" is obviously wrong for any reasonable definition of overlinking; if anything it would probably serve to prevent overlinking, as cases where links are used too many times can be reduced to a reasonable standard that serves readers, rather than to a "probably no one reads on a phone" standard that will obviously be reverted on sight as ridiculous. 60% of readers see articles in a context where they have no idea what's in a section they aren't reading and can't simply scroll up or ctrl+f to find it. (In fact, I'm unconvinced lower-technological-competence mobile readers know we have a search function -- we sure do a great job of hiding literally anything but "a small portion of the mainspace page being looked at" on Minerva.) Readers do not see articles as monographs where a missing link is 'just higher up'. The MOS should not have rules that make the encyclopedia worse. Vaticidalprophet 06:27, 15 June 2021 (UTC)
You're vastly overstating the rate at which links are hit, and understating the micro-disruptions to the reading process of continual changes in colour, not to mention the dilutionary effect to the linking system as a whole. And do you have any evidence that readers who parachute into one section would rather divert to another article rather than exploring the rest of the article they have arrived at? Tony (talk) 02:04, 16 June 2021 (UTC)
  • Oppose Since the proposal is too vague and if sections are short than it would be a sea of blue.  Spy-cicle💥  Talk? 21:48, 15 June 2021 (UTC)
    @Spy-cicle, could you say more about your "sea of blue" concern?
    We define, in this guideline, a "sea of blue" as meaning that people have linked two consecutive words to two different articles, like "[[Americans|American]] [[physician]]" (to give an example I saw earlier hours ago). A "sea of blue" is a problem because "American physician" looks like it's a single (two-word) link, but clicking on the first word doesn't take you to the same page as if you click on the second word.
    I'm guessing that your concern is not technically "a sea of blue" – there is nothing about a short section that makes the phrasing "[[Americans|American]] [[physician]]" more likely than "a [[physician]] in the [[Americans|US]]", which is not a sea of blue – so I'm wondering if what you mean is that you worry that editors will turn most of the words blue (which could be an aesthetic problem, but it's not a problem with figuring out where one link stops and another starts). WhatamIdoing (talk) 15:37, 18 June 2021 (UTC)
    @WhatamIdoing: To clarify I probably should not used the phrase sea of blue. Broadly speaking the more links we have on a page as the less links are used which is why we often should be using them sparingly and without having too many ones that are unnecessary as outlined at the start of MOS:OL. Moreover, linking terms more than once per section actually does more harm than good since increasing the abundance of links, removes the collective strength of all them in the first place, i.e. fewer readers actually using them, etc. Regards  Spy-cicle💥  Talk? 15:56, 19 June 2021 (UTC)
    @Spy-cicle, I wonder whether your concern is really "the more links we have on a page" (90 links in 1000 words is better than 100 links in 10,000 words) or is the concern about link density?
    Look at Breast cancer awareness. It's about 6,000 words. Looking at it today, I notice the following violations of this guideline:
    and there are probably more.  But as you look through it, does that article seem like it has has too many links overall?
    I don't see a proposal for linking every term, or even every 'key' term, in every section. I see a proposal for permitting editors to link a term as often as once per section, with the expectation that the actual result will be less than once per section. That is, the proposal is to move the rule from "once per article (links in the the lead don't count)" to "a maximum of once per section, if you believe once per article isn't enough for that specific link". WhatamIdoing (talk) 16:32, 20 June 2021 (UTC)
  • To clarify, I support the proposal of allowing highly relevant terms to be linked in every section. For many common words, the problem is not repeat linking, as they shouldn't be linked at all. Difficult technical terms should be linked reasonably often. No evidence has been given that allowing some extra links will create a sea of blue. —Kusma (talk) 22:46, 15 June 2021 (UTC)
    @Kusma: With the current wording ("generally"), as pointed out by others, we allow for exceptions which will should probably only really be used for much larger articles (also the fact this is a MOS guideline as opposed to a policy also allows for wriggle room). In regards to the example you pointed out I personally divided on whether pink ribbon should be repeated many times in the body. I can understand the organisations being repeated twice, given they are specific. Less of a need in my view for the third bullet point, and generally a more familiar concept to readers than specific organisations. Can understand the duplicate link for prostate cancer (suprised there is no prostate cancer awareness article though), less so for conflict of interest as a more familiar concept readers. Regards  Spy-cicle💥  Talk? 17:20, 20 June 2021 (UTC)
    @Spy-cicle, there are only two links to Pink ribbon in the body. (There is also one in the lead, and one in a caption, but MOS:DL exempts those.) Do you really think two links in such a long article is too much? (I sometimes wonder whether the articles should be merged, in which case there would be zero links.)
    I think that Prostate cancer and Conflict of interest are linked in both cases for the same reason: they're not super important concepts to the article, but you won't understand the specific point if you don't know what those words mean.
    Looking at MOS:DL, I don't think that the "generally" language allows for the exception that we've been asserting here. MOS:DL says "Generally, a link should appear only once in an article, but if helpful for readers, a link may be repeated in infoboxes, tables, image captions, footnotes, hatnotes, and at the first occurrence after the lead. That's "generally once per article, but as a special exception, you can have once in the lead and once in body and links outside the main body of the text". MOS:DL does not even hint at allowing two links to any page in the main body of the article. WhatamIdoing (talk) 17:43, 20 June 2021 (UTC)
    @WhatamIdoing: I mean looking at the example you provided you there I am fine a with linking to most of those terms twice. But adding one link per sections will inevitably lead to overlinking regardless of page size. Moreover, the "generally" and the fact this is guideline already allow flexiblity in my view. Regards  Spy-cicle💥  Talk? 17:49, 20 June 2021 (UTC)
    My point is that the "generally" exception is what's allowing the first link in the body. "Generally", the link in the lead is enough, "but" as a special, MOS-authorized exception to the general rule, you may also have a link at "the first occurrence after the lead". There is no "Generally, one link is enough, but you can have two in the body if you think it's important" in this sentence. If that's what you think it should say, then the existing wording is not doing what you want it to do.
    Nobody is actually proposing that everything be linked in every single section. The proposal is to set a maximum of once per section in the body, rather than the current maximum of once per article after the lead. WhatamIdoing (talk) 04:12, 21 June 2021 (UTC)
    The way that scripts such as User:Evad37/duplinks-alt are written (and used) is consistent with MOS:OL permitting one link in the body, full stop – not general flexibility. WhatamIdoing (talk) 04:33, 21 June 2021 (UTC)
    I see where you are coming from but at this point I think I am just refuting the central point, but I do thank you for your good faith attempts in the discussion. Regards  Spy-cicle💥  Talk? 12:41, 21 June 2021 (UTC)
    I'm looking for a guideline that means scripts would not flag additional links as errors. As long as software treats extra links as potential errors, even the most relevant and exceptionally-approved extra links will be under threat. So the guideline should be clearer that they are allowed, and do more to discourage overzealous script-assisted unlinking. (You pinged me instead of WAID about the cancer article, but she already responded anyway). —Kusma (talk) 06:41, 21 June 2021 (UTC)
    +1 to what @Kusma said here. These scripts are written to conform with the MOS as exactly as possible. If we don't want scripts flagging a second copy of a link a couple thousand words later, then we need the MOS to say that repeated links are okay. WhatamIdoing (talk) 03:25, 28 June 2021 (UTC)
  • Support to the general concept of relaxing stringency of 'one link per article' when warranted (especially in long articles). It comes up relatively frequently when a term is key two two separate sections. E.g. in Gene links to mutation are relevant in both the history section and the mutation section which may well be read separately. I appreciate that 'section' isn't strictly defined in all articles, but equally some articles currently transclude other articles, so 'once per article' isn't perfectly defined and both are less common than situations where multiple links would be useful in an article. I also think that the slippery slope of possible overlinking can be allayed via existing MOS:OVERLINK rather than an overly-strict one-per-article. T.Shafee(Evo&Evo)talk 06:27, 18 June 2021 (UTC)
    Long-term technical solution: the ability to hold a hotkey to have all linkable words shown as hyperlinks would be ideal. T.Shafee(Evo&Evo)talk 06:32, 18 June 2021 (UTC)
  • Oppose the original wording includes "generally" which allows for some level of common sense to be applied. We don't need some arbitrary "per-section basis" modification. The Rambling Man (Stay alert! Control the virus! Save lives!!!!) 15:41, 18 June 2021 (UTC)
OK, but every rule is vaguerized with that kind of language -- usually, with occasional exceptions, generally, consider, subject to common sense, etc. And of course, because if the rule said "without exception" that would be silly and it wouldn't have passed. But nobody much pays attention to them. Herostratus (talk) 00:08, 19 June 2021 (UTC)
So this proposal is a waste of time. It is already allowed to relink as appropriate, within the realms of common sense. I suppose perhaps if people lack such sense, everything needs to be prescribed. The Rambling Man (Stay alert! Control the virus! Save lives!!!!) 15:59, 19 June 2021 (UTC)
  • Support. I agree entirely with the points Kusma has made above. The biggest overlinking problem is in the form of links that shouldn't be included at all, rather than repeated links to relevant topics. Readers will often be thrust into the middle of an article by following a redirect. But also, visitors to an article who read it in its entirety from beginning to end are a minority - it's very common to jump around within an article, especially if you're looking for a particular piece of information. And also also, it's entirely possible for a reader to forget the identity of a person they read about 10 paragraphs ago. Editors should be encouraged to use common sense to lay links where they're likely to be useful. The current guidance of "generally no more than one link per article" is too strong of a statement to really reflect reality. Colin M (talk) 18:02, 18 June 2021 (UTC)
  • Support. On the merits, it's probably a net positive to the reader to allow writers to do this if they think it best. Give some credit to the writers to have a sense of what's best. Herostratus (talk) 00:08, 19 June 2021 (UTC)
  • Oppose It should be case-by-case based on length. There are times an article may lede with several short H2 sections (one paragraph each), and I would not want to see the same link repeated over and over in each of those sections. On the other hand, if I have scrolled through an H2 section that is two-some pages long (roughly) and now have come to a new H2 section, a repeat use of the link may be useful. It needs to be case-by-case, but definitely not requiring "once per section". The current allowance gives room for the situation I described. I would suggest that we can advice this repeat linking within sections deep in an article (not within visible range of the lede) that are known to be common targets for targetted linking from other pages (eg a common redirection term around video games, Crunch time (video gaming), drops into a deep section of Video game developer so repeating links there would make sense), but I would not do this for sections that are not used for such anchors. --Masem (t) 16:21, 19 June 2021 (UTC)
but definitely not requiring "once per section" But this is not at all what's being proposed. The proposed wording was: Generally, a link should appear only once in an article, but if helpful for readers, a link may be repeated [...] at the first occurrence in a section. Colin M (talk) 15:50, 20 June 2021 (UTC)
The wording is sufficiently vague that I can see that being gamed to be taken to read "Oh, that means every new section I can introduce a new fresh link." People will game this without something to speak to circumstances when a repeat link is appropriately. Some of the !votes above give the impression they would want to see a link per section, to that point. --Masem (t) 16:31, 20 June 2021 (UTC)
@Masem, for clarity, you are also opposed to the current rule, which does not normally permit a link to be repeated twice in the body of an article? (It's all well and good to say "but it's only a guideline and it says 'generally' [you know that word doesn't matter to a rules lawyer], and you can invoke IAR", but IMO we shouldn't be expecting to invoke IAR for every well-developed article.) WhatamIdoing (talk) 16:36, 20 June 2021 (UTC)
Recognizing that IAR allows for duplicate link allowances, I am not opposed to an improvement in wording to state that duplicate links in the body in latter sections are appropriate under circumstances, just that the suggested wording is too soft that this should be under very limited conditions. I don't think the current language is wrong, given that IAR exists. Its when we write it down that those that would game it and use WP P&G as red tape that we'd have problems. --Masem (t) 16:45, 20 June 2021 (UTC)
@Masem, it sounds like the practice that you'd like to see is:
  • current rules for linking in the lead
  • current rules for linking in the non-prose/body/text parts of an article (e.g., lists, tables, infoboxes, captions, hatnotes)
  • current rules about not linking too much in general
  • current rules about not linking unimportant terms
  • current rules about not linking the same term more than once in the same section
  • possibility of linking some terms more than once in the main prose/body/text of the article, particularly if the article/section is long
Does that sound about right?
I'm also guessing that you wouldn't object to removing the current rule that says the first occurrence after the lead should be linked (instead, let editors pick the best place, which will often but not always be the first occurrence). WhatamIdoing (talk) 17:51, 20 June 2021 (UTC)
  • Oppose Yes I have sometimes had to go search for a link that I remembered, and could even type it into the box, and maybe wish there were more links. I think, though that I agree with the US Supreme court: I know it when I see it even though it is hard to say in words. Larger articles can have more than one link, and I believe the current wording allows for that. Maybe a small change in the wording would be fine, but once per section is too much. Maybe every three or so screen-fulls in scrolling would not be too much. (For usual sized screen.) Gah4 (talk) 01:38, 20 June 2021 (UTC)
  • Oppose Agree with The Rambling Man, "as the original wording includes 'generally' which allows for some level of common sense to be applied." Overlinking is a FAR greater problem than underlinking, and surely all readers are familiar with the concept of copying and pasting a word or phrase into a search box. Edwardx (talk) 11:50, 27 June 2021 (UTC)
  • Support, but I would say something like "words or phrases in an article which have their own or very closely related Wikipedia article should generally be linked to the other Wikipedia page just once in an article, but may be linked to more than once in a long article or if it would be useful for the reader." BTW I don't think "conflict of interest" really needs a link on an encyclopedia page much more than the Pacific Ocean or buttocks. Oxybutinin on the other hand, might be useful to link more than once in a long technical article. Cheers, Facts707 (talk) 19:20, 10 August 2021 (UTC)

Discussion at Wikipedia:Village pump (proposals) § WikiProject links in navigational templates

 You are invited to join the discussion at Wikipedia:Village pump (proposals) § WikiProject links in navigational templates. --Trialpears (talk) 16:02, 8 July 2021 (UTC)

Multiple RfCs about overlinking

An editor has created multiple RfCs at several recent FIFA world cups about whether the nations should be linked: Talk:2002 FIFA World Cup#Inline link RfC Talk:2006 FIFA World Cup#Inline link RfC Talk:2010 FIFA World Cup#Inline link RfC Talk:2014 FIFA World Cup#Inline link RfC Talk:2022 FIFA World Cup#Inline link RfC. Since one was at Brazil and another Japan, the examples here might need to change if the decision is to link. Walter Görlitz (talk) 06:36, 25 July 2021 (UTC)

Three users (Rowsdower45, Tony1 and Walter Görlitz) have commented at all five of those RfCs; six more users (GiantSnowman, Herostratus, JBchrch, Lee Vilenski, Pincrete and Rubbish computer) have commented at one each. This risks a WP:OTHERPARENT situation; I'm not sure whether to close four of them, leaving Talk:2022 FIFA World Cup#Inline link RfC open (because that's both the oldest and the one with the most comments from different people) or to close all five and direct people to a central location (either here or WT:FOOTY). --Redrose64 🌹 (talk) 14:31, 25 July 2021 (UTC)
I closed them all - four on the grounds of WP:MULTI/WP:FORUMSHOP, one on the grounds that a broad-scope change should not be discussed at the talk page of an article, which is about as narrow a scope as you can get. --Redrose64 🌹 (talk) 15:19, 25 July 2021 (UTC)

I personally don't think these articles should be linked in any of these. I think the editor who opened these is forum shopping. Rubbish computer Ping me or leave a message on my talk page 14:32, 25 July 2021 (UTC)

  • Comment - my comments at the one RfC I !voted at would be the same for them all, and I support the discussions being closed apart from one. GiantSnowman 15:53, 25 July 2021 (UTC)
  • Thank you Redrose64. My actions were done late at night and I was initially wondering if I was tired or this really was a problem. Gladd to see my initial intuition was correct. I do think FOOTY would have been the right place to discuss it. Walter Görlitz (talk) 19:52, 25 July 2021 (UTC)

Seas of blue

Looking at MOS:SEAOFBLUE, there doesn't seem to be a clear rule on whether it's preferable to link per [[Riverside, California]] or Riverside, California for places that are suffixed by a state or county. A bit further down, it says

"For geographic places specified with the name of the larger territorial unit following a comma, generally do not link the larger unit. For example, avoid [[Buffalo, New York|Buffalo]], [[New York (state)|New York]]; instead use [[Buffalo, New York]]".

However, there may be instances in which it is useful to have the state or county following the place, and not everyone knows that Buffalo is in New York state (including self). (Also, if separated by a comma, it's not really a sea of blue.) Readers may like to read something about the state separately from the town name. I would like to see this spelt out to make it clearer whether it's supposed to be optional - IMO it could be left up to the discretion of the editor. We currently have a situation, e.g. where some places may be followed by a state within the same link in the infobox and others with a separate link for state (Berkeley, Gloucestershire vs Cheltenham; Newcastle, New South Wales vs Brewarrina; etc.). Laterthanyouthink (talk) 07:23, 3 August 2021 (UTC)

The larger territorial unit is presumbly already linked inside the article e.g. the Buffalo, New York page has NY state linked in its lead. As for display, it's up to editors whether to display the full Buffalo, New York or pipe to Buffalo if the state is redundant for a given context.—Bagumba (talk) 08:39, 3 August 2021 (UTC)
Yes, this is true, Bagumba - but I have come across different editors following different practices, and I have had both styles reverted on the basis of the same rule, doing opposite things, in the past! Hence thought it could do with spelling out. Laterthanyouthink (talk) 09:27, 3 August 2021 (UTC)
What did you have in mind? The example already suggests that separate city and state links are discouraged.—Bagumba (talk) 11:00, 3 August 2021 (UTC)
It's up to context: there are situations where either [[Buffalo, New York]] or [[Buffalo, New York|Buffalo]], New York or [[Buffalo, New York|Buffalo]], [[New York (state)|New York]] may be appropriate. Making mass changes or reverting simply "because MOS" is usually disruptive. Also worth pointing out that the buffalo rule was only added last month after this discussion. I don't think this was a good idea as the addition was made based on the strength of participating editors' personal convictions despite the fact that they were undermined by the little available data on reader usage. – Uanfala (talk) 13:50, 3 August 2021 (UTC)
Why are we going through all this again? --Redrose64 🌹 (talk) 14:41, 3 August 2021 (UTC)
Thanks, Bagumba, Uanfala, Redrose64. I raised it because I had not seen any earlier discussions, and didn't think it looked clear the way it is currently laid out and expressed. Perhaps the SEAOFBLUE notes could be put directly underneath the first point about piping, which demonstrates a place name, and spelt out that it's up to context how and what is displayed, so that it can be left up to individual editors' discretion? Laterthanyouthink (talk) 01:39, 4 August 2021 (UTC)
MOS:LINKSTYLE already shows the Riverside example, saying it can be directly linked (w/ California) or piped. MOS states no preference. However, MOS:SEAOFBLUE discourages side-by-side links, such as separately linking the city and state. Perhaps you could consider starting an WP:INFOPAGES if you wanted to see all the relevant guidance specifically for geo locations together.—Bagumba (talk) 02:25, 4 August 2021 (UTC)
Agree with Bagumba and Redrose. Tony (talk) 10:24, 5 August 2021 (UTC)

Help please on linking to a redirect rather than actual page

Hi all! What's the rule with this please? DeFacto has edited multiple pages changing the direct link to an article, to a redirect they created themselves, which then goes on to the actual article eg here and here. The article's name 'Treachery of the Blue Books', is the acceptable form; John Davies notes in The Welsh Academy Encyclopaedia of Wales that the name took hold of the public imagination to such an extent that ever since the report has been known by that name. Reference: see here. Isn't using another name to hide the true article name a form of censorship, and therefore in the spirit of Wikipedia, should be disallowed? Thanks! Llywelyn2000 (talk) 07:43, 30 August 2021 (UTC)

Llywelyn2000, WP:NOPIPE seems to be the appropriate 'rule' here. It is quite clear: "do not use a piped link where it is possible to use a redirected term that fits well within the scope of the text". However, if your question is more about your disagreement over the wording used in any specific article, then this probably isn't the appropriate talkpage to discuss that. -- DeFacto (talk). 08:48, 30 August 2021 (UTC)
So we should just always link directly to Treachery of the Blue Books. Looks very simple. Martinevans123 (talk) 08:58, 30 August 2021 (UTC)
I agree with Martinevans123- linking directly to the article name is better. Llywelyn2000 (talk) 18:39, 30 August 2021 (UTC)
Llywelyn2000, I thought you were asking what the rule was. -- DeFacto (talk). 18:47, 30 August 2021 (UTC)
Martinevans123, except that contravenes MOS:NOPIPE, doesn't it? -- DeFacto (talk). 18:47, 30 August 2021 (UTC)
I guess. But not the only consideration here. Martinevans123 (talk) 20:31, 30 August 2021 (UTC)
Agree. Linking to redirects is generally fine per MOS:NOPIPE. It's a content discussion for that specific article's talk page if the displayed text should be "Treachery of the Blue Books" or not; that is unrelated to MOS.—Bagumba (talk) 09:06, 30 August 2021 (UTC)
It's not a hard-and-fast "you must [not] do this" rule: it will depend upon the context. In some circumstances, a direct link to the article is best; in others, a link to the redirect is best - for instance, a redirect that is tagged {{R with possibilities}} or similar, which might one day become a full article in its own right. In the case of Treachery of the Blue Books that doesn't apply, because none of the inward redirects has any R templates at all. There are presently five redirects: 1847 Inquiry into the State of Education in Wales; Inquiry into the State of Education in Wales 1847; Reports of the commissioners of enquiry into the state of education in Wales; Reports of the Commissioners of Inquiry into the state of education in Wales; and Treason of the Blue Books, and there may be valid reasons to use one of these, particularly if one of these phrases is used by an authoritative source. --Redrose64 🌹 (talk) 09:04, 31 August 2021 (UTC)
Thanks Redrose64! So, if I understand correctly; direct link to article if that appears in the sources, otherwise a link to one of the 5 redirects (depending on the wording), depending on the exact wording used in the sources. So, if the sources use 'Treason of the Blue Books', then that is what Wikipedia should use in that part of the article (direct link). If the sources use the technical description 'Inquiry into the State of Education in Wales 1847', then that is what WP should use etc. Shows the importance of sources... and common sense! Best regards! Llywelyn2000 (talk) 06:30, 1 September 2021 (UTC)
Yep, that's pretty much it. --Redrose64 🌹 (talk) 07:39, 1 September 2021 (UTC)
Llywelyn2000, and like Bagumba said above, the phrase used in the article is decided on the article's talkpage - not here - and the syntax used in the link would naturally follow from that. -- DeFacto (talk). 07:53, 1 September 2021 (UTC)
Yes. The wording is an editorial decision that is outside the scope of the linking MOS. MOS:NOPIPE only advises not to pipe merely to avoid a redirect.—Bagumba (talk) 08:13, 1 September 2021 (UTC)

Linking to special pages, logs, etc.

Hi. I want to create an internal link to my block log for my bio page. How can I render this as a blue link? Piotr Jr. (talk) 18:49, 17 September 2021 (UTC)

Maybe ask your question (which has nothing to do with how links are styled) over at VPT, where just four minutes after you posted here, a user asked how to link to their log-list (not specifically block log). They got what looks like a good answer in 20 minutes. — JohnFromPinckney (talk / edits) 23:13, 17 September 2021 (UTC)
I posted there and was told it has something to do with how links are styled. Piotr Jr. (talk) 03:14, 21 September 2021 (UTC)
As JohnFromPinckney notes, your question is out of scope for this page which is for discussing improvements to the Manual of Style. Maybe, instead of starting a new section at VPT, you could have posted your question at the bottom of the existing thread. --Redrose64 🌹 (talk) 09:16, 21 September 2021 (UTC)

Linking to categories within an article

Is it OK to link to a category from an article's body? I'm looking at scholars of translation studies in the second paragraph of Skopos theory#Background. Largoplazo (talk) 10:00, 21 September 2021 (UTC)

It's unusual, but I think there are cases where it would be helpful to the reader. If you would normally want to link to List of foos but the article doesn't exist, linking instead to Category:Foos can be a reasonable fallback option. You could even use a cross-namespace redirect (with possibilities) for this purpose. e.g. right now List of Canadian provincial Acts redirects to Category:Canadian provincial legislation. In the future, someone might write a stand-alone list article at that name, but for now sending readers to the category is better than nothing. Colin M (talk) 23:34, 21 September 2021 (UTC)

OVERLINK and notable worldwide events

From a discussion on my talk page, would it be reasonable that within context of OVERLINK that there are certain worldwide events of significant scale that we should not link to when they're just name-dropped in an article?

This would be events like World War I/II, but possibly also going to things like the COVID-19 pandemic. Eg common we would see in a biography phrasing like "His parents immigrated to the United States from Eastern Europe in 1941 during World War II." and it can be argued that while adding WWII can bring significant in context, the link itself is unnecessary given that its a commonly known event to nearly all English readers. Or on many recent films "The film was delayed to (date) due to the COVID-19 pandemic." Of course, in these types of events, there are huge heirarchies of articles under them, and if a more specific link that was better context for the event. In the latter case, it would be better to have something like Impact of the COVID-19 pandemic on cinema as a target link, eg "The film was delayed to (date) due to the COVID-19 pandemic." or something like that, if it needs to be linked at all.

This is not to say that when WWII as a link itself can't be used when the war itself is the specific topic of relevant, eg History of Poland (1939–1945) better link World War II directly (among other links within that heirarchy).

And I would say this should avoid other major events that may be known worldwide but lack the worldwide significance, and thus may not be as readily known to all English speakers and that they should be linked - eg: Korean War, Vietnam War, September 11 attacks, Falklands War, French Revolution, Russian Revolution, etc., though editors should use reasonable discretion given context. --Masem (t) 18:51, 29 September 2021 (UTC)

Thank you, Masem. I routinely unlink World War I/II. Both of longstanding wide familiarity, and most unlikely to be clicked on. As for the COVID-19 pandemic, it is perhaps possible that one day it may fade from public memory as with the Spanish flu, but if so, that must be at least decades away, and I think it can be generally unlinked. Edwardx (talk) 11:24, 2 October 2021 (UTC)

Country in infobox

Hi, in Viktor Axelsen infobox, there were some linking to countries: Denmark and United Arab Emirates. Per MOS:LINKSTYLE and MOS:OVERLINK it is right or not? Stvbastian (talk) 14:43, 16 October 2021 (UTC)

Not. Removed. Nikkimaria (talk) 15:23, 16 October 2021 (UTC)

America linking

In the overlink section, there should be a section about America linking, Wikiedia's most infamous overlinks. Many pages have America links, including pages for people films, bands, companies, and even disambiguation pages (!) and user pages. --2A01:36D:1201:319:D7B:67F9:1B9B:5307 (talk) 06:14, 11 October 2021 (UTC)

We already cover overlinking to major countries in that section. {{u|Sdkb}} talk 01:06, 17 October 2021 (UTC)

Duplicate links and language templates

I had a minor discussion with User:FunkMonk on this, so figured I'd ask here. This might not be an interesting enough case to add to the guidance, but still interesting to see what people think. The current duplicate links guidance says:

Generally, a link should appear only once in an article but may be repeated, if helpful for readers, in infoboxes, tables, image captions, footnotes, hatnotes, and at the first occurrence after the lead.

What about when a link arises due to the expansion of a template - especially a template generally used in parentheses? I would personally argue that this is equivalent to the "footnote" case except with a classic parenthetical in-text footnote, and thus shouldn't "count" in general toward the duplicate links guideline. This is especially true if the link is relevant in the prose case and not-very-relevant in the templated case. The main example I'm thinking of here is the {{lang-XX}} family of templates that not only italicize a foreign phrase, but also include an explanatory link to the relevant language. For example, from the lede of the article Royal Spanish Academy:

The Royal Spanish Academy (Spanish: Real Academia Española, generally abbreviated as RAE) is Spain's official royal institution with a mission to ensure the stability of the Spanish language.

In my opinion, linking Spanish language in prose is "important" and useful here. The fact that "Real Academia Española" is in Spanish isn't as interesting, and the average reader's eyes will skip over the parenthetical part because that's what readers are trained to do. One solution would be to change the lang-es template to include |link="no" of course (France does this, for example) but I don't even think this should be necessary - the language link is harmless in the same way that repeated footnote links are harmless. Any thoughts on the matter? Does the duplicate link guidance extend even to links from expanded templates? If it does, how much latitude is there to IAR ignore this sometimes? SnowFire (talk) 21:28, 8 November 2021 (UTC)

I think this particular case warrants application of IAR; I don't think that's a common enough issue that it would require changes to the guidance. Nikkimaria (talk) 21:30, 8 November 2021 (UTC)
In my opinion, a link to the language is always WP:OVERLINK. How could the article on the Spanish language be of any help or provide any further useful information here? Further, the template {{Lang-es}} does have a parameter |links=no, so the above should be rendered "Spanish: Real Academia Española". -- Michael Bednarek (talk) 01:04, 9 November 2021 (UTC)
To be clear, the matter came up on an unrelated page, I just grabbed the RAE one at random as an example to not accidentally canvass. I'd argue that a link to the language is definitely relevant when it's in prose and specifically a matter of concern - if you read the linked article, the RAE is literally the institution that standardizes Peninsular Spanish. Anyway it sounds like you're more advocating that the {{lang-XX}} templates have their default changed to links=no, since I'll grant those links tend to be less relevant. SnowFire (talk) 03:49, 9 November 2021 (UTC)
I'm confused. I thought you were talking exactly about {{Lang-XX}}; that's what I see in your text, twice. As for the relevance of the RAE to the Spanish language: that's explained in the article's 1st sentence, so the link from the template is redundant. -- Michael Bednarek (talk) 06:28, 9 November 2021 (UTC)
Well, in the issue at concern, FunkMonk brought up duplinks, which was satisfied by removing the prose link later on that was relevant, i.e. the equivalent of keeping the first template-expanded link to Spanish but unlinking "Spanish language" from prose later. I'm arguing that we shouldn't even worry about duplinks at all and allow both links, although I agree with you that removing the link from the lang-XX template would be my second choice. SnowFire (talk) 13:39, 9 November 2021 (UTC)

Making it clear that linking within articles is permissible.

From MOS:SECTIONLINKS I surmise that it is okay to wikilink within an article. But this page doesn't clearly say that. Is there a place on this page to put that information? Butwhatdoiknow (talk) 17:25, 14 November 2021 (UTC)

That section seems somewhat out of step with general practice. The main situation in which one would want to link to a section within the same article is in the lead, but it can get confusing for readers if they don't know whether blue links in the lead are taking them to a section or another article. It's also generally redundant to the table of contents. For those reasons, I think we should focus on bringing the guideline in line with de facto consensus rather than trying to spell out a position not actually supported in practice. {{u|Sdkb}} talk 19:56, 14 November 2021 (UTC)
User:Sdkb: Sounds good. Would you please take a stab at that? Butwhatdoiknow (talk) 20:11, 14 November 2021 (UTC)
Done here. Cheers, {{u|Sdkb}} talk 20:33, 14 November 2021 (UTC)
Thanks. But, now that I've seen it, I wonder: what about linking to a re-direct in the same article? That wouldn't be redundant to the ToC. Butwhatdoiknow (talk) 01:22, 15 November 2021 (UTC)
Do you mean a circular redirect or something else? {{u|Sdkb}} talk 01:43, 15 November 2021 (UTC)
Let me try that again: what about linking to a shortcut in the same article? For example: putting a link to MOS:CIRCULAR in this article. That wouldn't be redundant to the ToC. Butwhatdoiknow (talk) 05:20, 15 November 2021 (UTC)
Oh, I see. This would be considered a page, not an article, since it's not part of the encyclopedia content itself. But yes, linking in places like {{shortcut}} is fine with me. {{u|Sdkb}} talk 07:11, 15 November 2021 (UTC)
Now I'm wondering about your view that "That section seems somewhat out of step with general practice." See comments below and, as an example, this recent article talk page edit saying "I don't see any problem with internal links to a part of the article further down." Let's continue this discussion below. (talk) 17:04, 15 November 2021 (UTC)

What is wrong with clicking on items in the TOC? Tony (talk) 22:33, 14 November 2021 (UTC)

Agreed. {{u|Sdkb}} talk 01:44, 15 November 2021 (UTC)
If a term on first mention in an article seems like something that would have its own article, but instead does have its own section in the article the reader is reading, then the TOC will serve the purpose of getting the user to that section only if the user memorized the TOC before starting to read the article and already knows that the term is there, right? Largoplazo (talk) 11:59, 15 November 2021 (UTC)
Agree with Largoplazo ("the TOC will serve the purpose ... only if the user memorized the TOC before starting to read the article"). Butwhatdoiknow (talk) 17:04, 15 November 2021 (UTC)
and ... only if the user realises that the section is relevant. I find internal links very useful if used judiciously. Readers who will never discover this discussion may do too. Certes (talk) 18:08, 15 November 2021 (UTC)
@Certes: Could you point to an instance of an internal link that you found helpful? {{u|Sdkb}} talk 18:11, 15 November 2021 (UTC)
I'm struggling to remember which ones I've actually clicked on but I'm thinking of cases like Animal (Major divisions in infobox) and Arsenal F.C. (the destruction of their South London stadium in lead). Internal links are also used for some reference styles, e.g. Wise, 1991 in American Civil War. Certes (talk) 18:26, 15 November 2021 (UTC)
The uses in Animal (an unusual circumstance in which the information can't fit in the infobox so there's a need to defer to the body) and American Civil War ({{Sfn}}s) seem fine to me. It's the type of circumstance in Arsenal F.C. (where the link is also broken, btw) that concern me. The reader should already expect that the stadium's destruction will be covered in the body, since leads are summaries of the body, and the fact that the year (1919) is given makes it pretty trivial to see that the "1886–1919" subsection of history in the Table of Contents is where to go for that information. Granted, there is a small benefit to having the link go directly to the section from the lead text. But if we introduce that as a new purpose for lead links, that creates a ton of problems. Should there be lots of links from the lead to every part of the body? How can readers then easily tell which topics in the lead have subarticles, and which links go to sections vs. articles? What should we do if there's both? It creates a lot of potential for confusion and makes it less clear which other related articles exist (which is crucial for reader navigation). Given that we already have the ToC for showing what exists where in the article and navigating to it, I very much question whether the redundant section linking is helpful. {{u|Sdkb}} talk 19:11, 15 November 2021 (UTC)
We certainly have some internal links which should be removed. I may have picked a bad example in Arsenal, as that link takes the reader to an anchor rather than the top of a section. I just picked the first results from a prefix:A search, so two out of three ain't bad. Certes (talk) 19:18, 15 November 2021 (UTC)
I question whether readers care much about easily finding out which topics in the lead have subarticles, and which links go to sections vs. articles. I would think they're much more concerned with easily finding out more about the subject of the linked text (wherever it may be). That said, I think we all agree that, if allowed, internal links should be used judiciously. You asked for examples, here's one: The link in the Domain Name System lede to Domain_Name_System#Resource_records. Butwhatdoiknow (talk) 19:36, 15 November 2021 (UTC)
Absolutely! I find such links quite helpful for long articles, especially ones with more complex structure. – Uanfala (talk) 00:41, 17 November 2021 (UTC)
I think it might be helpful to have broader input. I'll start an RfC below. {{u|Sdkb}} talk 04:49, 17 November 2021 (UTC)

Revert of 11:42, November 20, 2021

Regarding this revert, the reverting editor offers two concerns: First, "This is already covered in" MOS:SECTIONLINKS. Second, "it doesn't fit with the higher-level philosophy" at the MOS:BTW section on this page.
User:Sdkb: Regarding the first concern, the reverted text says why within-article links are appropriate and explains when they should be used. Where in the MOS:SECTIONLINKS (how) content can I find text that already covers why and when? - Butwhatdoiknow (talk) 22:31, 20 November 2021 (UTC)

@Butwhatdoiknow, if you'd like to add that sort of text to the MOS:SECTIONLINKS section, I'd be fine with that (ideally after the RfC above is closed, but I acknowledge the prevailing consensus). Just don't add another paragraph in the philosophy section, since when to selflink vs. not is about practical use, not overall linking philosophy. {{u|Sdkb}} talk 22:35, 20 November 2021 (UTC)
User:Sdkb: Correct me if I'm wrong, but it appears we now agree the reverted text is not redundant. Butwhatdoiknow (talk) 23:53, 20 November 2021 (UTC)

User:Sdkb: Turning to the second concern, you state that when to self-link is a practical rather than philosophical issue. What is your opinion regarding why self-links are appropriate - practical or philosophical? Butwhatdoiknow (talk) 23:53, 20 November 2021 (UTC)

The question of why self-links are appropriate or not is the subject of the RfC above. If there's agreement on that, it could be included briefly in MOS:SECTIONLINKS, but keep in mind that the MoS is primarily for explaining what to do; why to do it is often just WP:CREEP. {{u|Sdkb}} talk 00:52, 21 November 2021 (UTC)

BRFA courtesy notice

 You are invited to join the discussion at Wikipedia:Bots/Requests for approval/SdkbBot 2. {{u|Sdkb}} talk 21:20, 22 November 2021 (UTC)

The generally in MOS:REPEATLINK

I just looked and found the discussion from past May–June, where the general sentiment regarding repeated links seemed to be that some flexibility should be or is already allowed. However, some were concerned that the current wording didn't seem to convey this, and I concur. The current wording, "Generally, a link should appear only once in an article, but if helpful for readers, a link may be repeated in infoboxes, tables, image captions, footnotes, hatnotes, and at the first occurrence after the lead," reads to me as, "Generally, only once must apply. However, these six exceptions are allowed." This doesn't seem to reflect what most editors think the guideline should say or already does, which is that the entire once-per-article-plus-exceptions is but a general rule. I know this can get contentious very easily, so I'd like to suggest a very small modification for now, changing the phrasing to read:

Generally, a link should appear only once in an article but may be repeated, if helpful for readers, in infoboxes, tables, image captions, footnotes, hatnotes, and at the first occurrence after the lead.

I hope this will help convey that the entire sentence describes the general rule, which is what most editors think should be or already is. --Paul_012 (talk) 11:42, 3 October 2021 (UTC)

I can't tell the difference. Can you provide a specific scenario where the new wording would make a difference? Thanks.—Bagumba (talk) 16:12, 3 October 2021 (UTC)
Really? It moves "may be repeated" closer to only once, which I agree is a good idea, softening the rule and better reflecting what I think is the general attitude of experienced editors. It's not a matter of any "specific scenario". Support'. Johnbod (talk) 16:23, 3 October 2021 (UTC)
Really?: I did not interpret a difference, but did see the word order changes. No, it was not a joke. Yes, I might be the only one. Regards.—Bagumba (talk) 16:37, 3 October 2021 (UTC)
I do not see a difference either. MB 18:08, 3 October 2021 (UTC)
Ok, well you should have no objection to the change then. There is no change to the substantive meaning, just a useful shift in emphasis. Johnbod (talk) 18:19, 3 October 2021 (UTC)
I also support this change, for the reasons presented by Paul. Colin M (talk) 18:33, 3 October 2021 (UTC)
Article length should be taken into account. Long articles are less apt to be read top to bottom, and linking, say, every section should be allowed. That, at least, is my practice when I find such frequent links in a multi-thousand-word article when I'm copy editing. Dhtwiki (talk) 05:51, 4 October 2021 (UTC)
While I sympathise with this point of view, looking at the discussion from four months ago it seems unlikely that agreement on this will be easily reached, so I'm trying to aim for small steps right now. --Paul_012 (talk) 07:30, 4 October 2021 (UTC)

As it's been a week with input being either supportive or not seeing a difference (which I interpret as neutral to the change), I've implemented the edit. --Paul_012 (talk) 04:52, 10 October 2021 (UTC)

  • Support principle, as it's long past time we acknowledged that many readers aren't reading the full article (see research [1][2]) and give them links they are far more likely to find helpful than annoying. However, I'm not sure the suggested wording is sufficient to clarify the point that "generally" means repeated links are allowed, even if not in one of the specific exception categories. I'd prefer the wording Generally, a link should appear only once in an article, but it may be repeated if helpful for readers, such as in infoboxes, tables, image captions, footnotes, hatnotes, and at the first occurrence after the lead. Paul_012, could you link the April/May discussion so I can get a better sense of prior discussion on this? {{u|Sdkb}} talk 05:17, 11 October 2021 (UTC)

It doesn't allow an exception for a new section deep in a long article. It should. Herostratus (talk) 18:38, 7 December 2021 (UTC)

Revert

@Kwamikagami, you reverted me here, after I made an edit to follow up on this prior discussion. I agree it's contextual—would but generally not e.g. East Timor (with added "generally") be acceptable? Note also that the list is introduced Unless a term is particularly relevant to the context in the article, the following are usually not linked:, so there is already some wiggle room built in. {{u|Sdkb}} ' 05:20, 11 October 2021 (UTC)

We already link countries way too much, and your wording would just give people reason to think that's justified. It's relatively uncommon for a link to a country to contribute anything to an article, even a country like East Timor. Generally the question should be, Will this link provide the reader with additional relevant info, and improve understanding of the article? Usually the answer is 'no'. Most links on WP, actually, are completely useless and only obscure useful links. If it's just a matter of 'What's East Timor?', that's what the search menu is for. Per DICT, we shouldn't link to a term just because we suspect a reader might not know it, unless a search wouldn't easily find it. It's the same reason we shouldn't generally link to dates or languages. It's different with technical terms, where the reader won't be able to follow the text if they don't know the term, and it might not be obvious which article they should read to learn what it means. Or an abbreviation, because abbreviations are often ambiguous and a link would dab it. Or anywhere that a reader might want to go to explore the topic in greater depth. I could see in an article on a country, that we'd link to all neighboring countries, even major ones, because the reader might be interested in how they relate. Or we might want to link to the history or demographics section of a country article. But all of that could be argued on the basis of relevance, just as we might link to a date or a language if doing so would contribute something.
As one respondent put it, "Is the subject's interaction with the location vital to understanding either the country or the subject? This isn't a simple yes or no based only on the nation itself." — kwami (talk) 06:11, 11 October 2021 (UTC)
Courtesy pinging prior participants @Moxy, Walter Görlitz, Hemiauchenia, Khajidha, Tony1, Stepho-wrs, Johnbod, and Super Dromaeosaurus:. Cheers, {{u|Sdkb}} talk 18:40, 16 October 2021 (UTC)
I agree with Kwami. Tony (talk) 21:08, 16 October 2021 (UTC)
And I most certainly do not - see my comments in the previous discussion. I don't think that reached a consensus either way unfortunately. But his "Most links on WP, actually, are completely useless and only obscure useful links" is surely an extreme view - "many" yes, but not most. I don't actually think it matters much what the context is; I can't really think of any situation where I wouldn't link East Timor, yet I often go through articles removing many links. Links don't hold the reader up. Johnbod (talk) 21:40, 16 October 2021 (UTC)
For many English speakers (as their first or second language), I would think that the following countries would be well known - along with their popular (mis)conceptions:
  1. US - among world's largest economies, majority of English movies, large amount of literature, military dominance.
  2. England (but not Scotland, Ireland, Wales) - source of English language, source of much technology in the industrial revolution, previous economical strong, previous military dominance, previous English Emprie.
  3. China - among world's largest economies, everything is made there, sheer size, assumed by many to cover all of Asia.
  4. Japan - source of 2/3 of all cars, source of most electronics, "the enemy" in WWII.
  5. Russia - among world's largest economies, "the enemy" in the cold war, military dominance.
  1. Germany - fancy cars, "the enemy" in WW I and WW II.
  2. France - fancy cooking, fancy wine, shared history with England.
For many people, the other countries are unknown or only vaguely known unless you are next door or have a specific reason.
Even my own country of Australia is generally located as "elsewhere" (not a neighbour, not Europe, not America).
I'm quite well educated and have travelled a lot but I know that I would have trouble locating many countries in South America, Africa (beyond Egypt and South Africa), northern Europe or eastern Europe.
Quick quiz: do you known the locations of all the following beyond a vague Africa/Asia/etc ? Laos, Tanzania, Porto Rico, Transylvania, Uzbekistan, Liechtenstein, Tonga, Morocco.  Stepho  talk  00:59, 17 October 2021 (UTC)
I, for one, could show you all those places on a map (well, I could generally wave at the western Pacific for Tonga) and I've been to two of them (Puerto Rico and Liechtenstein). In addition, I could tell you that "Porto Rico" is a spelling not likely to have been published in English in the last 50, 75, 100 years.
I think one useful criterion is whether the mention is in a context where details of the country would likely be of interest rather than a digression. For example, "Jamie Performer went on to hold concerts in Laos, Tanzania, and Morocco before landing a standing engagement in Las Vegas" doesn't suggest that browsing over to the articles on Laos, Tanzania, and Morocco right now will add to user's understanding of Jamie Performer's career—but the nature of Las Vegas as a haven for long-term entertainer engagements could be useful. Largoplazo (talk) 01:18, 17 October 2021 (UTC)
Yes, I do know where they all are; but that's not the point. First, location is not everything. Second, in most cases readers don't want or need to know where a nation state is, and if one in a thousand want to drift there while they read an article, they can spend five seconds typing the name into the search box. Third, it's almost always better to give a more focused link than one to a whole country article: Sport in South Africa, Slavery in the United States, The economy of Switzerland—either by section link or offspring article. To avoid an easter-egg link (which almost no one will click on for its generality), better to link explicitly in a "See also" section at the bottom. Tony (talk) 05:08, 17 October 2021 (UTC)
Truly terrible advice, imo, especially the bit about cluttering up "See also" (I thought you were against doing this?). Obviously, if a more focused link is appropriate, use that. Our readers are used to floating over links they don't want to pursue, but the 'let-them-eat-cake' attitude of "they can spend five seconds typing the name into the search box" (and so lose the article they are reading, while hovering on a link will give them the start of the lead) is fundamentally not how we do things here. Johnbod (talk) 12:00, 17 October 2021 (UTC)
There seem to be strong differences of opinion here, so an RfC seems the best course of action. I've begun one below. Cheers, {{u|Sdkb}} talk 18:53, 17 October 2021 (UTC)

RfC: Linking non-major countries

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.
This RfC asks two questions. With respect to the first question – whether it is sometimes appropriate to add wikilinks to articles about less recognizable countries – there is a crystal-clear, essentially unanimous consensus that the answer is yes: articles about less recognizable countries (such as East Timor, the country that gave rise to this discussion) can and should be wikilinked if doing so would benefit the reader. With respect to the second question – whether the MOS's guidance on linking countries should be updated – there is no clear consensus at this time. Those arguing "no" (option C) argue that MOS:OVERLINK is already perfectly clear as written, while those arguing "yes" maintain that the disagreement giving rise to this discussion suggests that additional guidance is necessary. There's a lot of overlap between these positions, and many participants did not feel strongly about which answer was correct. At the end of the day, there's global consensus that linking non-major countries is acceptable under MOS:OVERLINK, and that consensus should be respected regardless of whether it's spelled out in an even more explicit way. If there's further confusion about this topic in the future, updating the guidance to expressly reflect this RfC's consensus would likely be uncontroversial at that time. (non-admin closure) Extraordinary Writ (talk) 07:20, 14 December 2021 (UTC)


When is it acceptable to wikilink countries, and what guidance should MOS:OVERLINK provide on this matter?

  • Option A: It is never acceptable to wikilink countries, but we should not change our guidance.
  • Option B: It is never acceptable to wikilink countries, and this should be noted in our guidance.
  • Option C: It is sometimes acceptable to wikilink less recognizable countries like East Timor, but we should not change our guidance.
  • Option D: It is sometimes acceptable to wikilink less recognizable countries like East Timor, and this should be noted in our guidance.

{{u|Sdkb}} talk 18:53, 17 October 2021 (UTC)

Background

This RfC arises out of a question that came up a little while back about whether or not to wikilink East Timor in a main page appearance. The current guidance is as follows (irrelevant bullets cut out):

Unless a term is particularly relevant to the context in the article, the following are usually not linked:

  • The names of subjects with which most readers will be at least somewhat familiar. This generally includes major examples of:
    • countries (e.g., Japan/Japanese, Brazil/Brazilian)

In subsequent discussions here and above, confusion has arisen both about what our guidance currently dictates and about what it ought to dictate. This RfC seeks to resolve those questions. {{u|Sdkb}} talk 18:53, 17 October 2021 (UTC)

Survey

  • Option D. Per WP:LINK, the purpose of wikilinks is to aid reader understanding of a topic by providing easy access to additional information related to it with which a reader might not be familiar. Linking major countries like the United States doesn't fit this purpose, but for a tiny country like East Timor, where most readers wouldn't be able to locate it on a map and the population is 10x less than some cities we'd link without second thought, it absolutely does. Our rules should reflect the editing practices that are best for readers, not the other way around.
    I believe that a plain reading of our current guidance permits wikilinking (note usually and major), but there is often confusion about this point, so I support adding East Timor as an example of a country that should generally be wikilinked to help clarify. This would only take a few words within the parenthetical, so there are no significant WP:CREEP concerns. I most strongly oppose option A, which would mean changing our rules but keeping guidance that contradicts them. {{u|Sdkb}} talk 18:53, 17 October 2021 (UTC)
  • Option D. Also change the policy so that links to countries in articles where they are obviously necessary are not removed. I explained some months ago there was one user removing links to Romania in the articles Parliament of Romania, Bucharest and a few others. Absurd. Super Ψ Dro 18:57, 17 October 2021 (UTC)
  • Option C, with no major objections to D. I do think it's already clear that links to countries that aren't major are permitted. I have yet to encounter debate over whether a given country is major/minor, but the added parenthetical won't much help if those are occurring. The other subcategories in the MOS list do not have examples of what doesn't count. Firefangledfeathers (talk) 19:24, 17 October 2021 (UTC)
    @Firefangledfeathers, I think the issue is that the guidance, while already clear, allows editors to gloss over caveat words, so it often gets simplified to "don't wikilink countries", which has become the de facto practice, even for countries as minor as East Timor as in the impetus discussion. Adding an example of a country to generally wikilink would help break us out of that false heuristic. {{u|Sdkb}} talk 19:50, 17 October 2021 (UTC)
    I can see that. A paraphrase of my comment might be: in my experience, there is not enough of a recurring issue that additional clarification is needed. I don't want my vote to be interpreted as opposition to D, but I can only !vote based on my experience. Do you have diffs handy of discussion that include misinterpretation of the current guideline? Firefangledfeathers (talk) 19:56, 17 October 2021 (UTC)
    I think the impetus discussion I linked just above is a pretty clear example; I don't have any others immediately on hand, but my sense is that it's a pretty frequent misinterpretation. {{u|Sdkb}} talk 20:08, 17 October 2021 (UTC)
    I don’t think the suggested change will affect the opposition in the discussion above. Firefangledfeathers (talk) 20:39, 17 October 2021 (UTC)
  • Option D. It seems pretty clear to me that WikiLinking obscure countries would be helpful for the majority of readers, and I think that it is already done to some extent today. Mover of molehills (talk) 20:40, 17 October 2021 (UTC)
  • Option C. As Firefangledfeathers said, MOS already covers it well enough. Unless we feel a need to use words of 1 syllable or less.  Stepho  talk  11:26, 18 October 2021 (UTC)
  • Option C / D. This is already clearly noted in the guideline, so no change is needed. pburka (talk) 17:42, 18 October 2021 (UTC)
  • Option C Tony (talk) 21:56, 18 October 2021 (UTC)
  • Option D per my previous comments. Johnbod (talk) 22:32, 18 October 2021 (UTC)
  • Option D but we need clearer guidance on what a "less recognizable" country is, or the same arguments will continue - Arjayay (talk) 08:15, 19 October 2021 (UTC)
  • Option D. Herostratus (talk) 06:15, 23 October 2021 (UTC)
  • Option D, although I think this is already implicit in This generally includes major examples of, and I tend to be wary of overly prescriptive guidance that seems to attempt to micromanage the entire project via the MOS. Archon 2488 (talk) 11:48, 26 October 2021 (UTC)
  • Option D as per Sdkb. LondonIP (talk) 00:50, 7 November 2021 (UTC)
  • Option C For the reasons given above for C&D Signed, I Am Chaos (talk) 06:45, 13 November 2021 (UTC)
  • Option C and to a lesser extent, Option D. It is the most user friendly option and consistent with the current state of the guideline. — Shibbolethink (♔ ♕) 03:28, 14 November 2021 (UTC)
  • Follow MOS:OVERLINK, namely avoid names of subjects with which most readers will be at least somewhat familiar. None of A–D exactly says this.—Bagumba (talk) 10:30, 14 November 2021 (UTC)
  • C. The guidelines already cover this, so revision isn't needed. If the wording can be improved without changing its actual meaning, then I have no objection to D.  — SMcCandlish ☏ ¢ 😼  15:46, 12 December 2021 (UTC)

Discussion

Courtesy pinging prior participants: @Moxy, Walter Görlitz, Hemiauchenia, Khajidha, Tony1, Stepho-wrs, Johnbod, Super Dromaeosaurus, and Largoplazo:. {{u|Sdkb}} talk 18:53, 17 October 2021 (UTC)

FWIW, I've always linked to every country, as long as it was only once per article. Example: I wouldn't repeatedly link to Canada, within the same article. GoodDay (talk) 19:04, 17 October 2021 (UTC)

The consensus not to link to major countries (like Canada) except in MOS:CONTEXTLINK situations seems pretty firmly established to me. We could add an option for "generally acceptable to wikilink all country names" if folks really wanted, but I think that'd have a snowball's chance of passing and might distract from the debate around A–D. {{u|Sdkb}} talk 19:12, 17 October 2021 (UTC)
I've no objections to allowing linking to all countries, past & present. But would advice that we re-word it as "sovereign state", to avoid the England/Scotland/Wales/Northern Ireland headaches. GoodDay (talk) 19:21, 17 October 2021 (UTC)
I don't agree with this; what matters is how likely readers are to benefit from the link, not the political status of the subject being linked to. So it is irrelevant whether it's a sovereign state, a sub-national state or territory, or a city (or has some other political status) – I think you cannot dispute that more English-speakers have heard of California, Wales, Ontario, Sydney, etc than have heard of Lesotho, Suriname, or Laos. Archon 2488 (talk) 11:52, 26 October 2021 (UTC)
But option D is already the status quo: the sentence that introduces the various bullets points (including the one for countries) is This generally includes major examples of:. It is clear that the rule for generally avoiding links to countries doesn't apply to non-major ones. – Uanfala (talk) 19:37, 17 October 2021 (UTC)
No thanks. That would be a retrograde step. Tony (talk) 21:55, 18 October 2021 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

RfC on self-linking within article prose

There is a consensus that self-links within prose should be allowed and that linking should be based on editorial discretion.
There is no consensus to actively encourage avoid using self-links. Conversely, there is no consensus to actively encourage using self-links in all articles. If that seems confusing, I would suggest comparing the situation to MOS:SEEALSO: Whether a link belongs in the "See also" section is ultimately a matter of editorial judgment and common sense. ... A "See also" section is not mandatory—some high-quality and comprehensive articles do not have one. Even though that guideline is not directly related to this question (and "See also" sections, in my experience, are much more common than self-links), the situation is somewhat similar.
All participants – even those on the opposing sides – seem to be promoting rather similar ideas, but with different emphasis. Nobody explicitly supported a blanket ban. Some have suggested that drafting the question collaboratively before launching an RFC might have led to different answers. That is something to consider next time. Politrukki (talk) 19:35, 2 January 2022 (UTC) (non-admin closure)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

In general, should articles be allowed to have self-links (wikilinks that go to sections or anchors elsewhere on the same page) in their prose? {{u|Sdkb}} talk 04:49, 17 November 2021 (UTC)

Background

The most common situation in which articles self-link in prose seems to be in the lead. For instance, the current revision of Arsenal F.C. links over the destruction of their South London stadium in the lead to an anchor in the history section where the destruction is discussed, and the current revision of Domain Name System links over Resource Records in the lead to the Resource Records section. This usage, while very uncommon, appears to be permitted by the language of MOS:SECTIONLINKS, which was added in 2010. Preliminary discussion above has revealed that there are differing views about whether it is desirable, something this RfC seeks to clarify. Please note that this only concerns links within article prose—other types of self-links, such as {{sfn}} references, links within infoboxes (e.g. at Animal), and self-links on project pages, are presumed to be permissible and are not under discussion here. The "in general" caveat also excludes unforeseen unusual situations in which we might want to make self-links. {{u|Sdkb}} talk 04:49, 17 November 2021 (UTC)

Survey re self-linking within article prose

  • No. We already have an element in the lead with wikilinks to the main parts of an article: the table of contents. It's where readers expect to find links jumping to sections in the body. Having self-links also appear in the lead prose is redundant to the table of contents. This wouldn't be so bad, except that it makes it impossible for readers to tell at a glance (i.e. without hovering over every link) whether a blue link in the lead goes to another article or just to a body section. This loss is crucial, since links are one of the main ways to navigate the encyclopedia. For instance, the article [[History of Foobar]] is likely to be found by reading the lead of [[Foobar]], seeing the blue link, and knowing that it reflects the existence of another article. If the lead of Foobar is instead filled with a bunch of self-links, readers suddenly will have no way to immediately see which topics mentioned in the lead have subtopic articles, since a blue link over history in the lead might just be going to [[Foobar#History]] instead. {{u|Sdkb}} talk 04:49, 17 November 2021 (UTC)
  • Yes There are good reasons why this sort of linking is very uncommon. But there are sometimes cases where it really makes sense to include a reference to an earlier or later section. We should trust editors to use their judgement on this, rather than having a procrustean blanket ban. The surprise factor of the reader hitting a self-link when they expect a link to another article can be mitigated by using the § symbol in the visible link text, such as is automatically inserted by {{Section link}}. Colin M (talk) 05:44, 17 November 2021 (UTC)
    @Colin M, the idea here isn't to issue a "blanket ban"; moreso just mild discouragement with plenty of caveats, more in line with de facto current practice.
    Regarding the § symbol, I don't really follow how that could be used in prose, as any self-link would presumably be piped. We're not discussing things like hatnotes here. Could you give an example of what you're talking about? {{u|Sdkb}} talk 07:28, 17 November 2021 (UTC)
    I'm thinking of self-references in prose, such as this example from the intro of Kolmogorov complexity: In particular, no program P computing a lower bound for each text's Kolmogorov complexity can return a value essentially larger than P's own length (see section § Chaitin's incompleteness theorem); Colin M (talk) 08:14, 17 November 2021 (UTC)
    Self-references like that are quite clunky; I'd ideally want to see them removed at the FA level. But they're not really what I'm targeting with this, so if you'd like to see them explicitly exempted, that'd be fine. {{u|Sdkb}} talk 19:02, 17 November 2021 (UTC)
    Sdkb, your the idea here isn't to issue a "blanket ban" is in conflict with the RfC as written; an outright ban is exactly what the RfC is about. If you mean for us to comment on discouraging or limiting or some other textual proposal, then please do rephrase the RfC to reflect that. — JohnFromPinckney (talk / edits) 12:33, 17 November 2021 (UTC)
    @JohnFromPinckney and Certes: What rephrasing would you like to see? I tried to be clear that the "in general" was intended as a wide carveout for any reasonable situations short of what you described as the "bizarre" example cases. {{u|Sdkb}} talk 13:52, 17 November 2021 (UTC)
    Well, for myself, I wouldn't like to see any rephrasing; I'm happy to see this thing snowdrift to a Yes = retain-the-status-quo close. But I guess one could try, "Should our guidance be changed to discourage the use of same-page links in the lead?" or something. That seems to be what you're actually concerned about, even if it varies from your actual request. But I agree with EEng below; it would have been good to collaboratively work out the RfC question in advance rather than giving a zero-seconds-long notice that you were going to start an RfC. — JohnFromPinckney (talk / edits) 00:12, 18 November 2021 (UTC)
  • Naturally yes, (preferably using {{section link}}) as it can be useful and appropriate in some cases (although not the bizarre case of ToC-obviating links in the lead suggested by Sdkb). I do not expect every bluelink to take me to a complete article somewhere else, nor do I expect a surfeit of links in the lead, but I do expect an appropriately helpful set of links to help me understand the prose I'm reading and provide details where I'm likely to need them. Outlawing same-page links would override our editorial discretion and keep us from providing in-page links in the (rare) cases where they're appropriate. — JohnFromPinckney (talk / edits) 12:33, 17 November 2021 (UTC)
  • Yes. I agree with the proposal to discourage rather than disallow, as implied in Background, but articles should still be allowed to have self-links where they are helpful. Perhaps we can agree some wording which encourages editors to think twice before doing so but continue if appropriate. Certes (talk) 12:46, 17 November 2021 (UTC)
  • Yes, and I'm getting pretty tired of half-baked RfCs opened by a single editor and launched unilaterally after only 48 hours of discussion, with no effort to frame the question jointly. An RfC should present a well though-out central issue, arrived at after thorough (not "preliminary") local discussion among watchers already interested in the issue, not present a giant dartboard for half-interested drive-by editors to aim random, uninformed comments at. It's telling that that almost immediately after this RfC was opened, it was suggested that it should have been phrased different. EEng 16:16, 17 November 2021 (UTC)
  • Yes. (1) The first sentence of the second paragraph of this page says "Appropriate links provide instant pathways to locations within and outside the project that can increase readers' understanding of the topic at hand." If the best target for an appropriate link is in the same article, why not use it? (2) As another editor said in the discussion before this RfC: "the TOC will serve the purpose ... only if the user memorized the TOC before starting to read the article." Butwhatdoiknow (talk) 16:44, 17 November 2021 (UTC)
  • Yes What is the problem that we are trying to solve here? That is, exactly why is it a Bad Thing for an article to link within itself? What harm does it do? As a general principle, a proposal for any new rule must first demonstrate that a problem exists with the the present situation that can only be addressed by the creation of a new rule. Or, what's broke that needs to be fixed? --Redrose64 🌹 (talk) 22:22, 17 November 2021 (UTC)
    @Redrose64: Did you read my !vote? You can agree or disagree, but I articulated a clear problem and rationale. {{u|Sdkb}} talk 22:35, 17 November 2021 (UTC)
    I did. The points you put forward concern the lead: but you apparently desire to forbid internal links everywhere, not just in the lead. --Redrose64 🌹 (talk) 23:31, 17 November 2021 (UTC)
    The confusion introduced over the purpose of links applies everywhere, not just the lead. I focused mostly on the lead because that's where self-linking is most common; I don't know of it happening elsewhere apart from the self-referential example Colin M gave above. What are the situations or examples in which you'd want to see self-linking from the body? {{u|Sdkb}} talk 23:36, 17 November 2021 (UTC)
  • Yes Very common in academic level articles and should be left to editor discretion at each article.Consider our data on what links are used it would seem that it could benefit the reader...... not seeing any harm in adding a navigational aid of this type.Moxy- 22:42, 20 November 2021 (UTC)
  • Yes - Obviously not every article should have these, and they really shouldn't be scattered willynilly, but on a case-by-case basis these can certainly be appropriate for the same reason ANY wikilink may be appropriate. If a reader comes across a term or concept they are unfamiliar with, they should be able to click the link and discover what they are missing. If we are writing an article and there is a term that can be reasonably expected to be unfamiliar to the reader, we should link it. If the location of the necessary information is elsewhere in the same article, and it is a sufficiently dense article such that the information is difficult to find, then an anchored link within the article is very appropriate. These are a lot of "if"s, but judging these things is best left up to the editors of individual articles, not a blanket ban. Fieari (talk) 07:23, 7 December 2021 (UTC)
    @Fieari: You (and several others above) seem to completely agree with me that in general self-links should not be used, with a few exceptions, so it's perplexing why you're !voting yes rather than no. There is no "blanket ban" here—please pay attention to the actual language of the question. {{u|Sdkb}} talk 16:46, 7 December 2021 (UTC)
    Sdkb, you're probably confused because they're (and we're) confused. The actual language of the question juxtaposes should articles be allowed (text connoting a blanket ban) with in general, which you seem to think means "but not really". So which is it? Do you want to allow or disallow this linking? And if your answer is, "it depends", or "in some cases", that why are we spending time discussing this proposal? What specific change in the MoS do you wish to see implemented. Write that, and then we can have a fruitful discussion. — JohnFromPinckney (talk / edits) 23:14, 7 December 2021 (UTC)
    Second that. Paradoctor (talk) 23:51, 7 December 2021 (UTC)
    I certainly don't read "in general" to clearly mean "not really", particularly not from the perspective of a wikilawyer coming in to an article and insisting to remove a useful anchor link because "the MOS says so!" Which is why I !voted yes. Including this text would indicate a blanket ban. There should not be a blanket ban. "It depends" and "in some cases" is exactly how it should be, and that's basically what we have already. Rules or policies put in place lead to robotic adherence. Lack of rules allows discretion, which is fine. Fieari (talk) 23:58, 7 December 2021 (UTC)
    The text that I'd want to see would be something along the lines of this:

    Links to a section of article from within the prose of that same article should typically be avoided (it's confusing for readers and redundant to the table of contents), but they may be included if judged particularly helpful. They may be used as needed in non-prose settings like infoboxes, references, and project pages.

    It's rather late to change the RfC question, but I'm curious to know if anyone here would support that. If so, perhaps I'll pursue it down the road, but if not it'd probably be better to let it drop and maybe just write an essay. {{u|Sdkb}} talk 18:19, 8 December 2021 (UTC)
    Still no. Either a link is helpful, or it is not. If it is helpful, then it should be added. If it is harmless, there is no reason to avoid it. Harmful links should be avoided anyway, so there is no need for WP:CREEP. There is no grounds on which to demand "particular" helpfulness. The essay suggestion I support, of course. Paradoctor (talk) 01:38, 9 December 2021 (UTC)
  • Yes I contest the idea that reader's heads will explode when a link goes to an anchor instead of another article. This is a solution in search of a problem. Paradoctor (talk) 08:23, 7 December 2021 (UTC)
  • Yes. Trust the editor to decide, that is how the Wikipedia was built. If it's egregiously stupid, someone will fix it eventually, and anyway it's not all that harmful. Herostratus (talk) 18:42, 7 December 2021 (UTC)
  • No. Generally, per everything Sdkb has said. I don't see any usefulness in these links – if the article's well structured, the reader will be able to find what they want in the Table of Contents – and it amounts to over-linking. JG66 (talk) 19:19, 8 December 2021 (UTC)
    That's a lot of if's. Why would we not want to help readers navigate main points? Why would the TOC contain terms. Have had a lot of ideas as of late that are detrimental to our readers.Moxy- 03:48, 9 December 2021 (UTC)
  • Yes. Relying on the TOC misses the critical point that users on mobiles [cell-phones] don't get a TOC. (A more useful proposal would require all such internal links to be prefixed with a pilcrow, so that it is obvious that they are going to a section in the same article. But that is another RFC.) --John Maynard Friedman (talk) 11:26, 20 December 2021 (UTC)
    You probably mean the silcrow: §. The pilcrow ¶ is for marking the start of a paragraph. Paradoctor (talk) 12:11, 20 December 2021 (UTC)
  • Obviously yes since it is already standard practice. The purpose of guidelines is to describe best practice not to try to legislate against practice with some new prescriptive idea, especially if the result would make it harder for the reader to find information, which would clearly be the result of this proposal, especially in long articles.  — SMcCandlish ☏ ¢ 😼  02:03, 21 December 2021 (UTC)
    Self-linking is very rare, not standard practice. I'd guess well over 90% of FAs do not self-link within article prose. {{u|Sdkb}} talk 02:21, 21 December 2021 (UTC)
    My guess is that well over 90% of all statistics are made up on the spot. But for discussion's sake, let's assume that number was real. That still leaves about 1 in 20 featured articles using them. That is not "very rare". If we considered self-links a side effect of making featured articles, they'd actually be common by your guess. Of course, "rarity" is not a standardized term, so the real question is whether current usage is relevant to the discussion. Paradoctor (talk) 05:44, 21 December 2021 (UTC)

Discussion re self-linking within article prose

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Geo place names

The guidline currently says:

  • For geographic places specified with the name of the larger territorial unit following a comma, generally do not link the larger unit. For example, avoid [[Buffalo, New York|Buffalo]], [[New York (state)|New York]] or [[Sydney]], [[New South Wales]]; instead use [[Buffalo, New York]] or [[Sydney]], New South Wales.

I have always applied this to "chains" of more than two places. I have run into a situation where an editor thinks that it only applies to the last element: [[Buffalo, New York|Buffalo]], [[New York (state)|New York]], U.S. instead of [[Buffalo, New York]], U.S.. I don't believe that adding the (unlinked) country nullifies the guideline as it applies to the other components of the name. Should we state this more explicitly in the guideline or at least give an example? MB 23:23, 21 January 2022 (UTC)

I've expanded the examples. Nikkimaria (talk) 23:46, 21 January 2022 (UTC)

Make guideline against linking to articles within Wikipedia that are not traditional content articles.

If you confused by what I'm talking about I'm going to give an example from the Donald Trump in popular culture article:

Trump has an article on the online encyclopedia Wikipedia. Editors of Wikipedia often have contentious discussions on what should be included in the article, and there are accusations of political bias among editors. The article has extended confirmed protection, so only certain editors (editors with at least 30 days of activity and 500 edits) can edit the article as well an unofficial editorial board.[75][76] The website "Loser.com" directs to Trump's Wikipedia article, and it is unclear who runs the site.[77][78]

In this excerpt notice how there is a link to the article Wikipedia:Protection policy. I think this type of link is inappropriate. There should be no linking to articles that are within Wikipedia, since these articles are not meant to be content articles, but pages to help editors. I have seen this a few times on Wikipedia and I don't believe it's addressed in guideline. It needs to be addressed and formally discouraged or banned. Iamreallygoodatcheckers (talk) 03:54, 10 February 2022 (UTC)

@Iamreallygoodatcheckers, it's addressed partially at WP:Manual of Style/Self-references to avoid, and was previously brought up here at Wikipedia_talk:Manual_of_Style/Linking/Archive_20#Linking_to_projectspace_in_articles. I agree with you that that should be prohibited, since linking from a mainspace article to another namespace page (non-mainspace pages are not articles) is essentially akin to adding an inline external link, and allowing ourselves to do it is basically giving ourselves preference over other entities that don't get inline external links. {{u|Sdkb}} talk 04:08, 10 February 2022 (UTC)

Use of word 'germane'

I linked the word germane to the wikitionary article on it. This was reverted by EEng, with the edit summary 'We assume our editors are basically literate' given as motivation. I think this is a fairly uncommon word. Per the Manual of Style itself, "Everyday words understood by most readers in context (e.g., education, violence, aircraft, river)" should not be linked, but germane is not such a word.

As I don't want to get into an edit war, I started this discussion page. Thoughts? TypistMonkey (talk) 02:18, 23 April 2022 (UTC)

I was also surprised that you linked it. I don't think of it as uncommon. Largoplazo (talk) 02:36, 23 April 2022 (UTC)
a) The advice in the MoS is for article content; the MoS itself is not a Wikipedia article. b) I agree that germane is not uncommon. c) It could be replaced with relevant. -- Michael Bednarek (talk) 02:40, 23 April 2022 (UTC)
I think it is uncommon (at least in British English), and therefore best avoided. There are plenty of alternatives for any given context. Johnbod (talk) 02:44, 23 April 2022 (UTC)
Using my benchmark of a high school freshman, I would say it isn't a common word to such a target reader. While I would probably not link it to the en.wp article because it's not likely to be directly relevant to the subject in question (because of the nature of the word), I don't think a link to wikt would be misplaced. -- Ohc revolution of our times 07:26, 23 April 2022 (UTC)
Following my reading of User:Michael Bednarek's comment above, I concluded replacing the word was the better solution, except I used "pertains to", which seemed a bit more suitable to me. I agreed: if a word is going to be much of a stumbling block, an interest in clarity ought to drive us to use more commonly recognized language rather than giving readers a homework assignment. Largoplazo (talk) 10:51, 23 April 2022 (UTC)
If you are going to link it, please don't use an external link: normal wikisyntax works perfectly well: [[wikt:germane#Adjective|germane]] → germane. --Redrose64 🌹 (talk) 12:30, 23 April 2022 (UTC)
Oh, whoops. Thanks for letting me know- I wasn't aware that was an option.
TypistMonkey (talk) 23:32, 23 April 2022 (UTC)
It's not nearly an uncommon enough word to need a Wiktionary link. If you think it is then a) perhaps you don't read enough other than newspapers and blogs, and b) you can try rewriting the sentence.  — SMcCandlish ☏ ¢ 😼  15:40, 25 May 2022 (UTC)

Disambiguation pages and draft articles

Hi. If non-existing articles have drafts, then visitors of article space see large editnotices. An example of what such notices look like:

Among other reasons, such editnotices are "in order to prevent the unnecessary creation of duplicates". As you know, some titles are ambiguous and therefore require a qualifier. Related to the above example, Luminary is the disambiguation page, while Luminary (podcast network) has a draft article associated with the ambiguous title. Editors who would like to - read; if it doesn't exist - create an article about the network, are much more likely to check the disambiguation page than they'll (correctly) guess/search+find the qualifier. My opinion is therefore that it makes sense to allow editors to link to such drafts on disambiguation pages. Currently, MOS:DRAFTNOLINK does no include an exception that allows such links. (Even though MediaWiki already uses {{Template:Draft at}} in article space.) Do you agree it makes sense to allow linking to drafts on disambiguation pages? I attempted to do so here, but my edit was reverted. --77.162.8.57 (talk) 10:36, 21 March 2022 (UTC)

Oppose - and not just because I was the editor who reverted the IPs addition.
On 29 July 2020 I started a discussion at Wikipedia:Edit filter noticeboard, which is archived here. This was taken up at Wikipedia:Edit filter/Requested but somehow, fell by the wayside. I have, however, been running daily searches for links to draft articles in mainspace.
I have not kept statistics, but I would estimate that over 60% of such additions are to drafts that have been rejected at WP:AFC, so are " .... not yet ready for the main article space, it is not in shape for ordinary readers ...." to quote Template:Uw-draft-link. The majority of such draft-links are added to to lists, typically, notable people, alumni or disambiguation pages. Assuming we are agreed that we do not want links to articles that have failed AFC, (whilst many of the others are ego/puffery pieces that clearly would fail AFC) I cannot see that there should be any exceptions. - Arjayay (talk) 15:41, 21 March 2022 (UTC)
Oppose per Arjayay.  — SMcCandlish ☏ ¢ 😼  15:46, 25 May 2022 (UTC)

Texas example should use hyperlinks grammatically

In the section "More words into a link", the guideline endorses the example:

In 1845, the Republic of Texas was annexed by the United States.

This is awkward because the hyperlink splits up the noun phrase "the Republic of Texas", taking only the word "Texas" from it. Reading the hyperlink alone results in a parse that conflicts with the sentence as a whole.

Including the entire noun phrase fixes this:

In 1845, the Republic of Texas was annexed by the United States.

Alternatively:

In 1845, Texas was annexed by the United States.

Jruderman (talk) 11:29, 30 May 2022 (UTC)

Wording with "Texas" or "Republic of Texas" is a hypothetical editorial decision, that is independent of the linking guideline. Assuming wording as "Republic of Texas" is the appropriate choice, the best practice for linking it would be to include all of "Republic of Texas was annexed" (but exclude leading "the"), as you suggest. —Bagumba (talk) 12:02, 30 May 2022 (UTC)

Consistency

There's a discussion at Talk:London#Linking of cities about whether we should be consistent and link to all cities mentioned in that article, or apply MOS:OVERLINK when an editor adds a link. I'd appreciate input, whether agreeing with me or correcting me, from editors more used to weighing such questions here than I am. NebY (talk) 20:55, 30 May 2022 (UTC)

Weighing MOS:SEAOFBLUE against desirable links

It's fairly common for articles to begin with a sentence of the form SUBJECT is an X Y, where X and Y are both linkable topics. That situation presents three possible paths, each with pros/cons:

  1. Rephrase the text to move one of the links later, e.g. SUBJECT is an X. It is a Y, and... Sometimes this works well, but often it results in clunky phrasing.
  2. Delink either X or Y in the first sentence. Sometimes it can be added farther down, but even when so, this typically makes it harder for readers to navigate to a defining contextual link.
  3. Keep both links in the first sentence, resulting in a sea of blue.

Let's assume for the sake of argument that there's no super elegant path available for options 1/2, as that'd be a cop out. I'm interested to hear from folks how you weigh the three options.

Personally, I tend to lean toward option 3, since research indicates readers don't tend to click links often, making the clunky phrasing of 1 a big downside, but when they do, position matters a ton, creating a downside to 2. Seas of blue are never ideal, but I'm not sure they actually harm enough to need to be avoided at all costs (something the guideline acknowledges with the when possible caveat), and I feel that they can be overpoliced because they're comparatively easy to identify. Thoughts? {{u|Sdkb}} talk 21:39, 5 April 2022 (UTC)

I concur with that entire post.  — SMcCandlish ☏ ¢ 😼  15:44, 25 May 2022 (UTC)
I'm interested in the general question here, rather than any specific example, but courtesy pinging @Dying and @Ravenpuff because you recently applied option 1 and 2 respectively to an upcoming TFA blurb. {{u|Sdkb}} talk 21:41, 5 April 2022 (UTC)
I generally aim for option 1, which I agree can result in a clunky first-sentence. If I can't make it work I usually just ignore it (option 3). In the interest of respecting the guideline I suppose I should use option 2 but there's also the fact that the next person to edit the article could have a better editorial sense than I do and can craft an option 1 sentence that reads perfectly well. Primergrey (talk) 14:01, 6 April 2022 (UTC)
Option 4 - "don't link either unless they really need it". Don't link "painter, scientist, historian, soldier" etc etc. Nor most countries. That removes 98% of the problem. The only problem is that eventually some idiot will come along and link "Italian" or "painter". Johnbod (talk) 15:03, 6 April 2022 (UTC)
@Johnbod, I'm certainly with you about the issues with overlinking, but MOS:CONTEXTLINK does allow a lot of links in the first sentence that you wouldn't normally want elsewhere, so I'm not sure it's quite 98%. For the sake of discussion here, let's assume the terms are uncommon enough to warrant links. {{u|Sdkb}} talk 17:22, 6 April 2022 (UTC)
Option 1 to rephrase. Something like Tom Brady's ...is an American football quarterback for the... could be re-written as ...is an American football player who is a quarterback for the... Every case is different, but here some might not know that a quarterback is a playing position.—Bagumba (talk) 17:38, 6 April 2022 (UTC)
Two links is not a sea of blue in my opinion. It also depends on how technical the article itself is. Walter Görlitz (talk) 16:38, 14 April 2022 (UTC)
By definition, two adjacent links make a sea of blue: "When possible, avoid placing links next to each other so that they look like a single link (a "sea of blue")". We see this a lot with city names (e.g., from El Paso, Texas) with just a black comma between. I generally just delink the state as being useless in that context (e.g. from El Paso, Texas). Too often, people throw in county and country, too, and link them all. Dicklyon (talk) 17:30, 10 June 2022 (UTC)

Easteregg links

The link, Phoenix, Arizona's Chinatown, is unclear because it could be read as either the Chinatown in Phoenix (as intended), or that "Arizona's Chinatown" is called Phoenix. I proposed fixing this by linking it as Phoenix, Arizona's Chinatown. Others have suggested that is an Easteregg link and should not be used. I think is sufficiently transparent. If you click on the link, would you be surprised/confused to be reading about the Chinatown in Phoenix, and not any Chinatown? This relates to a DYK hook and should be concise, not something unnaturally wordly. Thoughts? To avoid forking the discussion, Please comment at DYK. MB 20:35, 21 April 2022 (UTC)

The problem has nothing to do with it being a link, the displayed phrase is simply unclear. I can't see the context I'd need to reword it properly because you haven't shared the location, but suitable wording might be "the Chinatown of Phoenix, Arizona" or "Chinatown in Phoenix, Arizona". Largoplazo (talk) 22:13, 21 April 2022 (UTC)
Agreed.  — SMcCandlish ☏ ¢ 😼  15:41, 25 May 2022 (UTC)
How about Phoenix's Chinatown? Dicklyon (talk) 17:35, 10 June 2022 (UTC)

Reasons for repeat links

Hi there. There's a discussion on Talk:Christian Bale (edit | subject | history | links | watch | logs) about whether repeated linking is appropriate when the linked article's first and second mentions are contextually different and/or too far apart. I hope this page's watchers take interest in this discussion to help strengthen it. Thank you! KyleJoantalk 11:26, 9 July 2022 (UTC)

linking non-TLS Embassies and lots of everything that should be more secure

From Talk:List_of_diplomatic_missions_in_Ottawa

MANY years ago I noticed some wiki-bot(sic) tool Wikipedia was using to sanitize the pervasive use of INSECURE, privacy leaking, non-TLS links on Wikipedia. Seriously, what gives? Why do wiki pages such as 'List_of_diplomatic_missions_in_Ottawa', basically almost trick unsupecting users into clicking these links?

Isn't their wiki policy around this? There is no technical reason and this is extremely unethical. Fellow library users, wifi-hotspot fans, litterally multiple tech companies(and we know they log EVERYTHING) and whomever happens to be curious (i.e. listening in where the embassy site is housed(co-located) all get free open text viewing of this access of the embassy site. I did a quick perusal of policy, guidelines etc., but I'm at a loss and kinda shocked. I'm almost positive this is an already solved problem (automation fixes this easily), so why (and WHERE) are policy regarding this issue? 198.91.228.76 (talk) 00:31, 10 August 2022 (UTC)

Country links in country-related WP articles

An editor/reviewer on Aug. 9 twice reverted my WP link to the article "United States" in the first sentence of Languages of the United States. The link, he/she said, violates the WP:overlink policy of never linking "common terms." I'd say this is a "primary term," and every country-related WP article (ex., "Languages of...," "Demography of...," etc.) is linked to its country article upon first mention. The editor is unimpressed that 99.99% of country-specific articles in WP are already blue-linked this way; no, my link runs counter to the guidelines. I think the editor misreads "common term," and shouldn't the thousands of similar articles be de-linked to their country as well? Mason.Jones (talk) 14:32, 11 August 2022 (UTC)

The key for me is "the following are usually not linked" (emphasis mine). But what I understand you to mean by "primary link" certainly comes into play here. An article on a subtopic certainly should link its first mention of its supertopic. The content there is directly relevant information. It's not the same thing as linking it in the sentence "He spent the next five years in the United States before moving on to ...". Largoplazo (talk) 16:36, 11 August 2022 (UTC)
Yep, agree that articles should link to the next higher superordinate topic, as this is always highly relevant. Yes, even for targets like United States that all readers will have heard about. Links aren't done just to supply the meaning of potentially unknown terms, but also to situate topics in their proper context. It's for reasons like these that the same country is also linked from the introductory sentences of articles like U.S. state, History of the United States or President of the United States. Incidentally, the country link in Languages of the United States is that article's most clicked on link [3], so readers also clearly find it useful. Uanfala (talk) 19:16, 11 August 2022 (UTC)
@Mason.Jones Thanks for not notifying me of the discussion. BilCat (talk) 23:10, 11 August 2022 (UTC)
Glad we're all here now. On the issue itself, MOS:CONTEXTLINK would seem to support linking to the U.S. in that situation, based on what we've been told so far. {{u|Sdkb}} talk 00:11, 12 August 2022 (UTC)

Implementation in infoboxes

At (for example) the Queen Elizabeth II Diamond Jubilee Medal page's infobox, I've been trying to implement avoiding SEAOFBLUE. I've been reverted in that attempt. What have I done wrong? GoodDay (talk) 22:49, 14 August 2022 (UTC)

It looks that that got resolved. I have another issue. MOS:GEOLINK clearly says use Buffalo, New York instead of Buffalo, New York. Common sense says that this applies regardless of how the text is generated, it does not mean in prose only. Many infoboxes have a parameter like |location=, while others might have |city= and |state=, probably from before linking standards had evolved to where they are now. While changing these infobox templates would not be worth the effort, when I am editing one for another reason, I combine the info into one field to comply with MOS:GEOLINK (e.g. |city=Buffalo, New York and leave |state= blank. Occasionally I am reverted by someone who claims that each parameter must be used because they exist, and apparently, that an implied consensus existing from having legacy separate parameters overrides the expressed consensus of GEOLINK. See Chester School District as an example.
So I propose the addition of "This applies regardless of separate infobox parameters" at the end of the GEOLINK bullet. MB 21:11, 23 August 2022 (UTC)
Concur with MB. The fact that some infoboxes have a separate state parameter is not a magical loophole to ignore the guideline. They're just poorly coded templates.  — SMcCandlish ☏ ¢ 😼  06:12, 29 August 2022 (UTC)
GEOLINK is fine as is. Populating the |state= field does not require that it also be linked. I dont believe this one-off template requires any new wording in the MOS. Consider adding guidance like {{Infobox school district}}'s for |country=: "...see WP:OVERLINK.." or updating Wikipedia:WikiProject Schools/Article advice regarding GEOLINK.—Bagumba (talk) 15:11, 29 August 2022 (UTC)
@Bagumba, this does not apply to just {{infobox school district}}, there is {{infobox school}}, {{infobox university}}, and probably many more. (Although there are certainly others that use |location= or similar where this is no so much of a concern"). Adding "This applies regardless of separate infobox parameters" here covers them all in one place. MB 17:05, 29 August 2022 (UTC)
There's nothing in those infoboxes that says it needs to be linked, and community guidelines like GEOLINK would take precedence even if it did.—Bagumba (talk) 17:31, 29 August 2022 (UTC)
@Bagumba, if we don't have consensus to add a few words here (@SMcCandlish, were you concurring with the proposed MOS addition or just that "separate parameters are not a loophole"?) If we are going to clarify this in the templates instead, how does this sound:
This template has multiple parameters for individual parts of the geographic location. These parameters should not be used in a way that conflicts with MOS:GEOLINK. For example, use city = [[Buffalo, New York]] without using |state=, or use city=[[Buffalo, New York|Buffalo]] and |state=New York (unlinked).
I could put that into a template for easy reuse. MB 19:00, 1 September 2022 (UTC)
Sounds fine to me. I also don't have any real issue with adding the proposed wording to the guideline. MoS is long, but a few more words won't break it. Re: "There's nothing in those infoboxes that says it needs to be linked, and community guidelines like GEOLINK would take precedence even if it did." That truism has not, actually, prevented editwarring to keep over-linking the state, so something does need to be done here one way or another.  — SMcCandlish ☏ ¢ 😼  19:13, 1 September 2022 (UTC)
I stand by the truism, but it's ok if consensus is reached without unanimity. Regards.—Bagumba (talk) 02:32, 2 September 2022 (UTC)
If a template has separate parameters for the different subdivisions, then it's best practice to continue using them as they allow for better structuring of the information and would enable decisions about formatting (including linking) to be taken at the level of the template. Infoboxes don't need to be exempt from the MOS, but given that they present information in a different way from prose, it can't be taken for granted that everything that applies for one context will automatically be relevant for the other. And, given that style guidelines are rules of thumb that have at best rough consensus among the community and that, per policy, they allow for exceptions, I don't think enforcing them with zeal in the face of local opposition is a particularly constructive activity. Uanfala (talk) 12:07, 2 September 2022 (UTC)
While WP:LOCALCONSENSUS doesn't necessarily supersede an MOS, one shouldn't edit war over this also. A discussion like this that attempts to determine if a local exemption has broad support or not is the right direction.—Bagumba (talk) 13:46, 2 September 2022 (UTC)
Doesn't this lead to another inconsistency? In one article we'd have city = Buffalo, New York and not show state on its own line, in another using the same infobox template we'd have city = San Diego, state = California. NebY (talk) 11:21, 6 September 2022 (UTC)
That level of "consistency" is not required. Per MOS:OL some people only specify the city for places like Paris, or New York City. In the example template here, the template (infobox school district) outputs city and state on the same line, so the city and state are on one line regardless. Also, many of these templates have an alternate |location= that is entirely free-form, where the editor can specify line breaks and format the address/location any way they wanted. MB 14:01, 6 September 2022 (UTC)
I agree with MB. I see overlinking in inboxes far too often. An audit to fix this problem would be welcome. Particularly annoying are the sport-team infoboxes that have a field for "Current season". This will be carelessly filled with a plain year rather than waiting until a specific article is created, usually after the current year. Tony (talk) 09:44, 7 September 2022 (UTC)
The discussion above seems to imply something that's perhaps more extensive than the proposed wording, which might mean the effect's unclear or simply that I'm being slow. To take four examples, please, is the linking in the infoboxes of Dulwich, Hanley, Bradford Grammar School and Indian Military Academy already in breach of MOS:GEOLINK or would this change make it so? NebY (talk) 11:38, 10 September 2022 (UTC)
The (rather misleadingly named) guideline at GEOLINK applies to run-on links of the type London, England, and so is effectively a special case of MOS:SEAOFBLUE. This is not applicable to articles like Dulwich or Hanley, where the links will all appear on separate lines. MOS:OL may be relevant in those cases, but it's worth noting that almost all articles about populated places will have all the administrative divisions (incl. the country) linked in their infoboxes. Bradford Grammar School and Indian Military Academy, where the link appears on a single line, are closer to what is being discussed here. Uanfala (talk) 14:24, 10 September 2022 (UTC)
Ah, thanks. So "This applies regardless of separate infobox parameters" would be contrary to {{Infobox university/doc}}, the example in {{Infobox school district/doc}} and the UK example in {{infobox school/doc}}, and frequent if not common use of those templates. To avoid conflict or disruption, the template documentation should be changed if MOS:GEOLINK is changed, and it might be good to have some degree of acceptance from some of the editors who have a particular interest in school and university articles. As yet, I don't see any notification of this discussion on either the template talk pages or at Wikipedia talk:WikiProject Schools and Wikipedia talk:WikiProject Higher education. NebY (talk) 16:07, 10 September 2022 (UTC)
I have been quite busy recently but just posted a notification of this discussion to both those projects. MB 00:24, 25 September 2022 (UTC)

Is the only use of the city, state, and country parameters (at least in the university infobox, the only one that edit) to display a location in the rendered infobox? I ask because I am wary of making that assumption without checking to see if those parameters are used in other ways, some of which I assume have been subsumed into Wikidata. If that is the only use of those (and any other similar) parameters, I would be happy to help adjust the template's documentation to reflect this project-wide consensus. (If it matters, this is not the decision that I would make nor the one I like; I merely acknowledge that local consensus or practice cannot override project-wide consensus.) ElKevbo (talk) 02:58, 25 September 2022 (UTC)

Summary

As requested, I posted notification at Wikipedia talk:WikiProject Schools and Wikipedia talk:WikiProject Higher education and in four weeks, there has been one additional comment.

In summary, I see the following:

  • MB agrees project guideline takes precedence, add statement for additional clarity
  • SMcCandlish agrees project guideline takes precedence, add statement for additional clarity
  • Bagumba – agrees project guideline takes precedence, MOS change not necessary but OK with consensus
  • NebY – only noted that template documentation should be updated and projects notified
  • Tony - agrees project guideline takes precedence, add statement for additional clarity
  • ElKevbo – agrees projects guideline takes precedence
  • Uanfala – allow use of template parameters as they are coded

There was one objection, one that did not express an opinion, and the other five expressing varying levels of support. I believe this is a consensus to add the statement "This applies regardless of separate infobox parameters". MB 21:28, 22 October 2022 (UTC)

There was broad agreement that the project guidelines apply to infoboxes, but that doesn't translate into support for including a mention of this fairly obvious consideration into the wording of the particular subguideline at MOS:GEOLINK. If I'm not mistaken, only two editors (MB and SMcCandlish) supported such an inclusion. In case I haven't made that clear earlier: I oppose that. It is unnecessary (why should this one linking rule among dozens of others on this page mention infoboxes?), it's likely to lead to overgeneralisation and misapplications (see NebY's comment), and will probably encourage the misuse of template parameters (when the correct solution would be to change the template itself). – Uanfala (talk) 10:14, 23 October 2022 (UTC)
I don't support this change and I do remain unhappy about making a widely applicable change with so little directly positive input. The MOS, like WP policy and guidelines, has largely developed as a codification of established practice, and this is not in accord with practice with some heavily used infobox templates. Also, it's a bit of a pink flag when an editor seeks to change the MOS to back their disputed edits but doesn't bring the editors with whom they're in disagreement into the discussion, eg in this case I see MB raised Chester School District as an example but didn't ping the editor with whom they were in a slow edit-war, @Alansohn: NebY (talk) 15:01, 23 October 2022 (UTC)
There is no policy change here. The guidance in GEOLINK of how to link places is exactly the same as before. Wikipedia:Manual of Style/Linking is a project guideline. Contrary documentation in project templates is only a WP:LOCALCONSENSUS. This statement is just to clarify that the guideline has precedence, even in infoboxes. As far as individual notifications, I did not notify anyone. I am aware that similar changes have been reverted before, this was just the first time I decided not to ignore and just move on. I was not about to go through thousands of my past edits looking for more examples. That would have been a huge waste of time and would have biased the discussion in one direction. MB 17:51, 23 October 2022 (UTC)
Yep. This is basically an RfC without an RfC tag, and the consensus is nearly unanimous.  — SMcCandlish ☏ ¢ 😼  21:37, 23 October 2022 (UTC)

PIPESTYLE

MOS:PIPESTYLE probably should mention that the "derived names" format should not be used with piped links (for example, although [[Class (set theory)|class]]es works, [[Class (set theory)|classes]] is preferred). — Mikhail Ryazanov (talk) 17:45, 15 October 2022 (UTC)

Does this come up so often (and fixes for it revert-warred so often) we really need a rule about it?  — SMcCandlish ☏ ¢ 😼  21:40, 23 October 2022 (UTC)
Hmm... I stumble upon such formatting from time to time and convert it to the more readable form, but don't remember explicit edit-warring about it (although I don't know whether other users might be doing the opposite without attracting my attention). So, probably, this indeed doesn't need a special rule. I guess, if I encounter some opposition, I can then just refer to this discussion. — Mikhail Ryazanov (talk) 14:56, 24 October 2022 (UTC)

Is this considered over linking?

If you have an article about various chart achievements for a particular music chart, is it over linking to repeatedly link songs and artists who appear in multiple sections? Or are those instances fine? On the List of Billboard Hot 100 chart achievements and milestones page, almost every table or section is completely linked but then there are places in prose and/or some tables where Taylor Swift for eg. is linked and then places where she is not. I'm working on fixing up the World Digital Song Sales page and I'm a bit confused on where to remove links and where to add them (obvious over linking in the same table aside, which I am getting to slowly). Or is it that a reader should be able to click anywhere in any table and be taken to the relevant song or artist page regardless? -- Carlobunnie (talk) 17:21, 18 October 2022 (UTC)

Articles that are collections of tables are different from prose articles. In a prose article, a reader is more likely to proceed through the article from top to bottom, and overlinking repeated terms becomes more of a concern. In articles with a bunch of tables, of which your example is a good one, a reader is more likely to jump straight to their table of interest and so having links there is a appropriate. FWIW, tables are explicitly listed as an exception in MOS:LINKONCE. Regards, Orange Suede Sofa (talk) 00:45, 19 October 2022 (UTC)
Ah, idk how I missed the tables mention in the section that you linked so thank you for bringing it to my attention and for clearing up my confusion. -- Carlobunnie (talk) 04:41, 19 October 2022 (UTC)
I concur with Orange Suede Sofa.  — SMcCandlish ☏ ¢ 😼  20:00, 10 November 2022 (UTC)

Linking sports seasons pages

Is it permissible to link a year as part of a date within a table? e.g. a sports season page such as linking "28 October 2022" if it were within a table of a golfer's win list? Obviously per WP:DATELINK, linking year articles should be avoided e.g. 2022, however is it possible to link the PGA Tour season within the date as similar to the [[1787 in science|1787]] example given in the policy outline. Jimmymci234 (talk) 19:26, 28 October 2022 (UTC)

Sounds like it would be an egg. Nikkimaria (talk) 03:32, 29 October 2022 (UTC)
It could be like the [[1787 in science|1787]] example, if someone would reasonably expect such a link there. But given the full-date example above, "28 October 2022", I doubt this is the case. It would probably help to see the table. In general, though, I can't think of any other sports articles doing this, except in tables that have something like a season column, with links to season articles that are obviously links to season articles.  — SMcCandlish ☏ ¢ 😼  20:03, 10 November 2022 (UTC)
Examples at Séamus Power (golfer)#Professional wins (7) or Tony Finau#Professional wins (5). Tewapack (talk) 21:46, 10 November 2022 (UTC)

Discussion at Talk:Miss Universe 2022 § MOS:OVERLINK does not apply?

 You are invited to join the discussion at Talk:Miss Universe 2022 § MOS:OVERLINK does not apply?. ☆ Bri (talk) 17:35, 14 November 2022 (UTC)

Wikipedia talk:Manual of Style/Lead section has an RFC

Wikipedia talk:Manual of Style/Lead section has an RFC for possible consensus. A discussion is taking place. If you would like to participate in the discussion, you are invited to add your comments on the discussion page. Thank you. InfiniteNexus (talk) 07:03, 30 November 2022 (UTC)

RfC about hatnote links to Simple English Wikipedia

Should hatnote links to Simple English Wikipedia be discouraged? fgnievinski (talk) 20:09, 8 October 2022 (UTC)

For an example, see Ring (mathematics): "For a simpler explanation, see simple:Ring (mathematics)."
For an alternative, see Template:Introductory article and Template:See introduction.
For background, see Simple English Wikipedia and Wikipedia:Simple English Wikipedia.
  • Yes, because English Wikipedia editors do not necessarily have control over the content of Simple English Wikipedia. Furthermore, simple English is defined by the use of a reduced vocabulary, not introductory explanations. fgnievinski (talk) 20:19, 8 October 2022 (UTC)
  • Meh. I agree that links for introductory information on an article's main subject to anywhere outside this Wikipedia is problematic for the reasons given (controlled by their own editors with differing goals). If this were routine, we'd need to take steps to encourage a better solution. But I haven't seen this happen. Where it does, it's likely just a short-term solution until our article is improved (incorporating our own version of the external information, for instance). Having the link might even encourage our editors to fix our article. Unless it's somehow a thousand times more prevalent than I can see, I don't think it's broken enough to warrant making yet another MOS rule. --A D Monroe III(talk) 21:10, 19 October 2022 (UTC)
  • Yes. We recently deleted the template that was used for this purpose, {{Simple}}, and this RfC seems like another way of having that same conversation. Copying my rationale from there:

    There are established pathways for interproject linking, and those should be used instead of arbitrarily choosing some articles to attach this to. Because Simple English is about language choice, not complexity, it's not clear what those would be — {{Introductory article}} exists for inherently complex subjects, and articles that needlessly use advanced language can be tagged with {{Technical}}. In both cases, we can handle the issue here on English Wikipedia without pushing readers to a different language version.

    Responding to A D Monroe III's creep concerns above, a "yes" result here doesn't necessarily mean that we need to expand the MoS if this isn't a recurring issue; editors can just link back to this RfC anytime it comes up. {{u|Sdkb}} talk 21:36, 19 October 2022 (UTC)
  • Yes, I would discourage inter-language links in hatnotes in general. SWinxy (talk) 22:19, 19 October 2022 (UTC)
  • Yes, especially since the template for it was nuked at TfD. It's kind of a combination of WP:FAITACCOMPLI and WP:FORUMSHOP to continue the practice sub rosa after a clear consensus against it already.  — SMcCandlish ☏ ¢ 😼  21:39, 23 October 2022 (UTC)
  • Yes—too much clutter already. Plus lack of control over another site's content. Tony (talk) 04:13, 6 December 2022 (UTC)

NOFORCELINK

Hi! I've recently had this brought up at Wikipedia : Featured article candidates/2022 World Snooker Championship/archive1 about the MOS. I understand the need for making articles readable from any background, and that you shouldn't need to have an understanding of the topic to understand an article. However, where is the limit on explaining a term that is common in the subject matter, where a link is not enough?

The above example is regarding the use of break in an article on snooker, which is about as common a jargon term as there is, and I've also seen an argument about this on the terms of "yellow card" and "goal" in football, but similar terms might be for a "shooter" in video gaming or a broadcast for television shows.

My big issue with explaining in prose a common term that takes a while to explain is that it doesn't make the article have better prose, and slows down reading for an informed reader. I understand the current wording is intentionally quite vague, but perhaps if we add something to the effect of it being unnecessary to overly explain a term if it interupts the flow of an article.

Let me know your thoughts. It's not that I object so much to this term, but I can see it being applied to every article (or theoretically could be) and wouldn't improve how articles are written. Lee Vilenski (talk • contribs) 18:30, 2 January 2023 (UTC)

Hi, I'm the editor who brought this up in Lee's FAC. I refer to MOS:NOFORCELINK: "Do use a link wherever appropriate, but as far as possible do not force a reader to use that link to understand the sentence. The text needs to make sense to readers who cannot follow links" fairly frequently. Today in the middle of this with examples; here on 27/12; several times here two weeks ago. It applies to me too, eg I have had to explain in several articles that "Pikemen were equipped with pikes: long wooden shafts tipped with steel points. Pikes as issued in both armies were 18-foot-long (5.5 m), but on the march they were commonly cut down to a more wieldy 15 feet (4.6 m) or so" despite this being proverbially "as plain as a pikestaff". You may have gathered that my concern is largely around FAC where there is a firm requirement to adhere to the MoS and, less rigidly, even more important than usual that information is actually conveyed to a reader. Gog the Mild (talk) 19:49, 2 January 2023 (UTC)
Thanks for expanding on this Gog, and apologies for the lack of ping, I had intended too! Whilst this is something that seems to come up at FAC, the MOS applies to all articles, so I would like a larger input on exactly how this should be handled. For the same reason we don't need to cite that the sea is blue, I don't think terms that most readers of that topic would be familiar with are similarly needing explaination. Maybe I'm over sensitive, but explaining every term on every article wouldn't be a great read in my opinion. Lee Vilenski (talk • contribs) 21:18, 2 January 2023 (UTC)
No worries. Hopefully you can see that I am neither picking on you nor sports articles. I'm picking on everyone and getting picked on myself. There is some slack in "as far as possible", but not a lot. Personally, if a term is passably clear from context or one can ignore it and still follow the main thread, then I tend to give it a pass. Naughty of me I suppose. But if it really needs to be understood to follow the thread of the article then at FAC I ask that the MoS be met; to me this doesn't seem to unreasonable. Yes, the policy applies to all articles, but I suspect that little reversion on the grounds of NOFORCELINK happens in practice.
I tend to have a stock of ready explainers, in some cases as footnotes, for things I use repeatedly in my articles. Eg pike, musket, brigade, body armour, dragoons etc. Or for age of sail articles striking colours, rating ships by number of guns, Old Style and New Style dates, signal frigate etc. All complete with a handy source.
While many readers of your fine cue sport articles will be aficionados, many will not, and will not be able to follow the article if they don't understand what a "break" is. And are you aware that if they do follow the link, the first thing they will read is "Typically describes the first shot in most types of billiards games" which is not what you mean at all. Gog the Mild (talk) 21:42, 2 January 2023 (UTC)
In terms of it being a bad link, I agree, I've long since thought I should expand century break to be about breaks with a section on century breaks, which might fix that. Lee Vilenski (talk • contribs) 14:26, 3 January 2023 (UTC)
Until then, would it be helpful to set an anchor at the second definition of break? It seems {{cuegloss}} works with anchors and that hovering over such a link in an article would show the anchored definition. NebY (talk) 15:36, 3 January 2023 (UTC)
One point of MOS:LINK is:

Do not unnecessarily make a reader chase links. If a technical term can be meaningfully explained with very few words or with a footnote, this may be preferred to a link.

At a minimum, I expect technical sports terms to be linked. I think it's a balance between being helpful to a reader who knows nothing about the sport versus a casual follower finding it verbose or even patronizing. If I was reviewing a page for a sport I knew little about, I'd make the point about making it accessible to non-fans, while deferring—within reason—to the sport experts on how to balance it. AGF on both sides is essential.—Bagumba (talk) 14:02, 5 January 2023 (UTC)
The bit about footnotes in the passage from MOS:LINK quoted above was only added a few hours ago [4]. Is this newfound preference for footnotes a good idea? Clicking on subscript letters to get to the notes is a form of chasing links (though less inconvenient than when the link target is a different page). – Uanfala (talk) 14:35, 5 January 2023 (UTC)
Indeed, today's change is from if a highly technical term can be simply explained (my emphasis), and the removal of "highly" does not seem to have been discussed. For example, the article about a snooker championship also links the technical terms frame, foul shot, black ball, doubling, clearance, fluke, in-off, free ball and session to spare. Would it benefit readers of this article to have each of these terms explained or footnoted, or would the clutter worsen the readability of the article unnecessarily, given its likely readership? Are we also imposing too much on editors and throwing away the great advantages of a hyperlinked encyclopedia? NebY (talk) 16:38, 5 January 2023 (UTC)
Indeed - this is very much the issue at hand (although black ball might be hard to explain... It's a ball... that is black.) Overly detailed or complicated things I do feel the benefit in clarifying, however, I agree with the above that we do benefit from not interupting prose to explain every term we see, that we do have links to articles (or in this place a glossary) that explain the term better.
Of course, taking into account the fact that the above term isn't well described on the target article (I'm working on that). Lee Vilenski (talk • contribs) 16:44, 5 January 2023 (UTC)
Given that the MOS change to If a technical term can be explained with very few words or with a short footnote, this is preferred to a mere link represents a major shift away from hyperlinking and is not in accord with MOS:UL "In general, links should be created for ... Articles explaining words of technical terms, jargon or slang expressions or phrases—but you could also provide a concise definition instead of or in addition to a link", I've reverted. NebY (talk) 17:06, 5 January 2023 (UTC)

I ran into some similar issues last year when doing GA reviews for several articles on ice hockey players. I know little about hockey and had questions about quite a few terms: natural hat-trick, taxi squad, top pair, and several others. I could see that explaining everything inline would be disruptive for readers familiar with the sport. I asked for links, and as I recall everything I had a question about ended up being linked. I don't know that we can define clearly the boundaries between terms that need inline explanations, those that need footnotes, and those that just need to be linked -- it's going to be different for different readers. "Pike" is completely obvious to me, though I think it should be linked; a century break is also obvious (I grew up watching Pot Black), but I can see it means no more than "taxi squad" to someone who's never played or watched snooker. I don't know how to codify this, but I think if knowledgeable readers feel that the flow is being broken up too much by inline explanations, non-knowledgeable readers should accept that, and be willing to accept links instead, or, if absolutely necessary, an explanatory note. The canonical example of this sort of problem is mathematics, where Laplace transform, a not very obscure mathematical tool, could not possibly be explained to someone with a math background in a single article. Mike Christie (talk - contribs - library) 21:30, 8 January 2023 (UTC)

I think for me, it's a no brainer to link the term, we should attempt to do that as much as possible, but if we should explain items in prose. Lee Vilenski (talk • contribs) 21:40, 8 January 2023 (UTC)
I think that the way the MoS is currently phrased pretty much reflects what Mike says above.
" "Pike" is completely obvious to me". Well there you go: the Macedonian pike of Alexander's army, the Scottish pike of Medieval schiltrons, the European pike of the Thirty Years' War and later, the half-pike of the Napoleonic Wars, and the half pike still carried by the 3rd US Infantry Regiment on ceremonial occasions are all different beasts. (Let not even think about Chinese, Japanese and Indian variants down the centuries.) A bare "pike" just doesn't cut it at FA. Any more than "break" does, even if we know we're talking about a cue sport. We're an encyclopedia. We're supposed to explain things. Surely that's what "prose is ... of a professional standard" means? Gog the Mild (talk) 22:02, 8 January 2023 (UTC)
I actually read Christie's reply as being that we should link terms, but that it breaks up the prose too much to explain something then we should settle for links. Let me give you an example of a term that is easy to link, but difficult to explain from my own area of expertise - free ball. If we had to explain that in prose (it would take me a good paragraph to explain) it would completely kill the flow of an article. A break is a series of points made in one visit to the table, which is easier to explain, but it doesn't actually matter if you know what it is, just that it is a thing, and there's a prize for it. I think for me, we need some better ground rules as to when it makes sense to explain in prose (of footnotes) and when to only leave a suitable link. Lee Vilenski (talk • contribs) 22:27, 8 January 2023 (UTC)
Mike's if knowledgeable readers feel that the flow is being broken up too much by inline explanations, non-knowledgeable readers should accept that, and be willing to accept links instead, or, if absolutely necessary, an explanatory note is what I'd thought one of Wikipedia's guiding principles as a wide-ranging hyperlinked encyclopedia. To find an example, I looked in Wikipedia:Featured articles for a baseball article (never played it, never watched a game, no interest) and found 2004 World Series. This non-knowledgeable reader isn't familiar with third straight wild card team to win, swept the Cardinals in four games, bench-clearing brawl, RBI single, single, double, top of the fourth, skip off the lip of the infield, right field foul pole, left-center field bullpen, a single to right field that scored Mueller, earned run.... I made guesses about some and kept going, much as I would if reading a newspaper article.
Now imagine the average US reader's reaction if bench-clearing brawl[a] and top of the eighth[b] were distractingly marked as if someone who wanted further details about the brawl could find them in the footnotes[c], just like in a book, and started clicking. They might not agree that it was even a Good article and suggest, in nontechnical and profane terms, that it be reviewed. NebY (talk) 01:10, 9 January 2023 (UTC)

Notes

  1. ^ where the brawlers sweep everyone into a pile at the end of the bench
  2. ^ the high end
  3. ^ footnotes appear at the bottom and are signified with little letters

MOS:GEOLINK

I don't see why country shouldn't be linked. Its generally educationally helpful to the reader to state the country especially if its name/type of state is different from what is present in that geography today. Comrade-yutyo (talk) 12:05, 28 January 2023 (UTC)

Take a look at more of the same guidelines: MOS:SEAOFBLUE, MOS:OVERLINKING. Linking every word that happens to refer to a subject about which Wikipedia has an article would be overwhelming. The guidance, as I see it, is to consider whether the information at each other article would provide additional value in the context of the article being read. If an article, for example, says that in 1974 a band toured in cities such as Chicago and Seattle, the reader (a) probably has at least heard of these places before, (b) probably isn't driven to want to know more about those cities just because they are the ones that happen to have been mentioned among the perhaps 50 cities they visited that year, and (c) doesn't need to know anything about those cities to understand the article being read.
But suppose an article says that a particular actor spent the earlier part of their career in Chicago. Possibly knowing something about Chicago, specifically, might be relevant to gaining an appreciation of that actor's experience. But it's less likely that higher level information about Illinois or the United States or North America is relevant to the reader at that point; and if that reader goes to Chicago and, from there, does feel driven to learn more about Illinois, then Chicago has a link for that.
It's a matter of striking a balance between being useful and creating unnecessary distraction. Largoplazo (talk) 12:45, 28 January 2023 (UTC)

an example of sea of blue?

OK so I'm having a bit of an issue with Chiswick Chap here in [5] - he says that linking Wagner in that sentence is not an excessive amount of linking, despite the fact that this was a sequence of three nouns linked. Now it's a sequence of four nouns where "opera" isn't linked, but I still don't quite see the point. Can someone else assess this? --Joy (talk) 11:44, 15 February 2023 (UTC)

Um, thanks for pinging and discussing. There is now no issue: there never was any overlinking as alleged, and there is now no sea-of-blue - it was never anything much to write home about (2 items) but now they are as Joy says no longer contiguous, so the issue, such as it was, is resolved already. Chiswick Chap (talk) 11:47, 15 February 2023 (UTC)
For what it's worth, I almost always tell people to not have links to two different items right next to each other unless it is completely unavoidable. Lee Vilenski (talk • contribs) 12:37, 15 February 2023 (UTC)
Many thanks. Chiswick Chap (talk) 12:47, 15 February 2023 (UTC)
So then you agree that linking all of Siegfried, Wagner and Der Ring des Nibelungen is excessive? I mean, I'm just not seeing the logic here. The addition of the word opera doesn't actually help in my mind. The average reader of Dwarves in Middle-earth is so unaware of any of these that they would honestly benefit from having each of those links, but they don't need a link to opera, because people interested in Tolkien are already acutely aware of theatical art forms, yet don't know who Wagner is? --Joy (talk) 22:07, 15 February 2023 (UTC)
celebrating Siegfried in Wagner's opera Der Ring des Nibelungen seems fine to me. I might spell out "Richard Wagner", as if he's unfamiliar enough to link, he's also unfamiliar enough we should give his full name.
If you want a true sea of blue example, that's here. {{u|Sdkb}} talk 22:19, 15 February 2023 (UTC)
Looking at the example stated, there's no reason not to link Wagner, unless he's already been linked. Lee Vilenski (talk • contribs) 20:56, 16 February 2023 (UTC)
All of those are good links.
Note that we don't (primarily) provide links because the subject is unfamiliar. Links are navigational elements, not an alternative to a dictionary. Add links whenever you think someone might want to read that article, too, even if you're certain that the readers will recognize the name. WhatamIdoing (talk) 18:13, 24 February 2023 (UTC)

Is there a consensus that we should not give links to the article about the nomination of US federal judges?

User_talk:Snickers2686#John_G._Koeltl. Apokrif (talk) 23:08, 7 March 2023 (UTC).

As a non-American, the American judicial system and how it nominates judges is a bit of an unknown to me. If I was interested in John G. Koeltl for some reason then the nomination process might be useful. However, his nomination didn't seem controversial. In which, case, the link to United States federal judge at the top of that article would cover what I would likely be looking for (ie, what makes a US federal judge different from any other type of judge). The link to United States federal judge#Appointments (via a piped link of "nominated") does not provide anything above and beyond the first link. If his nomination was controversial then the link to United States federal judge#Appointments would be more useful.  Stepho  talk  23:30, 7 March 2023 (UTC)
The anchored link spares the reader the pain of going back to the beginning of the article, looking back for a potential link to an article which might have a relevant section. Apokrif (talk) 00:02, 8 March 2023 (UTC)
No, there is no such consensus, but I don't imagine that something so specific is what anybody is claiming. Are you going to make us guess where this link is and which link it is, or make us search through your contribution history to find it? Largoplazo (talk) 00:37, 8 March 2023 (UTC)
The reason for this question is in Snickers2686's talk page section linked above. Apokrif (talk) 01:08, 8 March 2023 (UTC)
I see nothing there that implies the existence of a consensus such as the one you've asked about in the heading of this section. So whatever's going on in that conversation, I think you're misunderstanding the point. Largoplazo (talk) 01:31, 8 March 2023 (UTC)
I asked whether there was a consensus, I did not claim there was one :-) Apokrif (talk) 01:35, 8 March 2023 (UTC)
I don't have a problem with that wikilink. I think a reader could plausibly click on it and find it useful. (If you're looking for instances of overlinking, there are certainly far worse ones in the same article. e.g. "He studied history at..."). One mild negative is that a reader who sees that link might assume it leads to an article (or section) about Koeltl's nomination, rather than the nomination process for federal judges in general. Colin M (talk) 00:44, 8 March 2023 (UTC)
@Colin M, the guideline says the opposite: that when you link "nominated" people will assume it's about nomination in general, and not about Koeltl's nomination specifically. I was copyediting that sub-section just hours ago; it's MOS:MORELINKWORDS. If you think the section is wrong (I think the section might overstate the risk of potential confusion, but I like its suggested approach), then we should consider revisions. WhatamIdoing (talk) 01:39, 8 March 2023 (UTC)
I don't disagree with that guideline. It's true that, if the wikilinked text was instead "Koeltl was nominated" it would create a much stronger presumption that the link leads to information about Koeltl's nomination specifically. I still think there's (unavoidable) potential for confusion. But I don't think it outweighs the utility of the link. Colin M (talk) 13:22, 9 March 2023 (UTC)

Reading difficulty

@Masem, do you disagree with both of the sentences you removed?

  • Please do not add or remove links on aesthetic grounds or your personal perception that it's hard to read.
  • Research indicates that links to other articles do not make it harder to read the text.[1]

I kind of doubt that you really want people to add and remove links on aesthetic groups or their personal perception that it's hard (or easy) to read. WhatamIdoing (talk) 23:48, 10 February 2023 (UTC)

I am disagreeing with the use of the graduate thesis to justify these. We would not use that for mainspace content, so we shouldn't be using it to justify policy/MOS issues. Masem (t) 01:16, 11 February 2023 (UTC)
Indeed. Having done my PhD on eye movement in text reading, I can say that the methodology suggested in that abstract raises questions. Tony (talk) 03:58, 11 February 2023 (UTC)
I thought it was better to provide some evidence that the assertion is true, although that's not a requirement for policies and guidelines. If you'd like, we could cite a different source, such as
But perhaps of more immediate relevance, @Masem, if your problem was with citing a source, would you please go put back the first sentence, which has nothing to do with the cited source? It's better not to remove two sentences plus one source, when you only dislike the source. WhatamIdoing (talk) 06:56, 12 February 2023 (UTC)
I don't have time to look through that thoroughly, and I don't want to appear carping; but in the abstract the claim that we move our eyes to bring new information into our fovea looks as though she hasn't read Kevin O'Regan's work on probability and peripheral vision. I don't buy any of these references. And let's not forget that we don't want to dilute the linking system, so we ration it. Go to fr.WP or any other language WP to see the patchy ugliness of unrestrained linking. Tony (talk) 08:28, 12 February 2023 (UTC)
@Tony1, I agree that links should focus on relevant points. What I want to stop is editors claiming, with zero evidence for this position and some evidence against this position, that it's hard to read linked text. I can't imagine you want editors to make up "alternative facts" about visual difficulties. They need to stop saying "It's hard to read" and start saying "This isn't allwiki, and we need to link the important things without distracting people with less-relevant links". I'm convinced that won't happen until we provide a little education here, by directly stating that our choice is about relevance, not about reading difficulties. WhatamIdoing (talk) 18:09, 24 February 2023 (UTC)
@Tony1, I see that you removed this advice: "Please do not add or remove links on aesthetic grounds or your personal perception that it's hard to read linked text."
Do you think editors actually should be adding and removing links on aesthetic grounds? WhatamIdoing (talk) 23:42, 16 March 2023 (UTC)
No, although a sea of blue certainly does look messy. The problem with your theory is that it's too vague; I can see it being used to beat editors over the skull for raising the bar for what is worth linking and what is not. And you have no consensus here for adding that. Tony (talk) 03:37, 17 March 2023 (UTC)
I'd like to have something in this guideline to point at, when an editor explicitly says that there should be a different number of links because
  1. they personally believe that a different amount of blue text would look better, or
  2. they personally believe that a different amount of blue text would be easier to read.
I don't know if it's possible to be very specific – MOS:LEADLINK says that "Too many links can make the lead hard to read" without giving any advice on what constitutes "too many", which is already vague – but I'm willing to be as specific as we can be. What would you tell such an editor? WhatamIdoing (talk) 06:28, 17 March 2023 (UTC)
That's because the need for links that a reader is likely to click on varies a lot. Is your agenda to end up with more links than we already have? That is my suspicion. Tony (talk) 07:29, 17 March 2023 (UTC)

References

  1. ^ Martikainen, Hanna (2018-10-31). "Mind the Links! How Hyperlinks Influence Online Reading and Navigation : An Eye Movement Study". {{cite journal}}: Cite journal requires |journal= (help) Graduate thesis in psychology, University of Turku, 2017

DL, sections, and mobile readers

There is a consensus for the change. I've edited the MoS per the wording suggested. Mike Christie (talk - contribs - library) 13:05, 13 June 2023 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

I'd like to challenge the assertion in MOS:DL read from top to bottom like a regular article.
I saw a lot of people state during the recent Vector 2022 RFCs that they don't read Wikipedia articles the way they read a novel, and that they come here for some specific information on a topic rather than reading the whole article top to bottom. On mobile, sections are collapsed by default, so it's actually super inconvenient to encounter some term that has been linked once somewhere above in the article, and now I have to expand all the sections individually and hunt for where it was initially mentioned and linked.
Another negative experience I've had due to application of this policy is encountering a term in the context where it first becomes interesting enough to check out on its own, but again it's been mentioned in passing somewhere above, and I have to go find the first instance to follow the wikilink. I'll also add that when a list of related topics are introduced somewhere in an article, and exactly one isn't wikilinked because it's been mentioned already, it looks unattractive.
What I'm wondering is if we can loosen the guidance in the DL section to ...may be repeated if helpful for readers, such as in infoboxes, tables, image captions, footnotes, hatnotes, and at the first occurrence after the lead in a section. This should help accomodate mobile readers, as well as desktop readers who click anchor links from the TOC instead of reading the entire article. Folly Mox (talk) 14:33, 13 May 2023 (UTC)
  • I'd be happy relaxing the rule from once per article to once per section. --Redrose64 🌹 (talk) 20:57, 13 May 2023 (UTC)
  • Support for the reasons given by Folly Mox. - Butwhatdoiknow (talk) 23:38, 13 May 2023 (UTC)
  • Support also, because one-per-section is closer to actual practice, especially in long and complex articles.  — SMcCandlish ☏ ¢ 😼  10:02, 5 June 2023 (UTC)
  • Support for the reasons outlined by Folly Mox, which basically make the links more accessible. It would also help in the specific case when someone's name is first mentioned in passing and then in the next section that person is the subject. Mujinga (talk) 10:21, 5 June 2023 (UTC)
  • Comment This could mean that we end up having a terms linked multiple times in a long article (7 or 8 times, possibly more) and possibly fairly close together, if at the end of one section and the top of another. I know common sense would dictate against that, but common sense isn't that common and often doesn't make sense. - SchroCat (talk) 10:30, 5 June 2023 (UTC)
  • (edit conflict) Comment -- common sense should prevail; I think the current practice permits more than one mention in the main body in long and complex articles, for instance in a detailed composer article it seems logical to allow the same term to be linked in both "life" and "works" sections, OTOH it seems excessive to link a term in one short section and then again in the next section. Cheers, Ian Rose (talk) 10:35, 5 June 2023 (UTC)
  • Support per above – good change IMO. J947 † edits 10:42, 5 June 2023 (UTC)
  • Support. Seems a sensible change, and of particular benefit in longer articles. – Michael Aurel (talk) 11:41, 5 June 2023 (UTC)
  • Support per above.  Stepho  talk  11:42, 5 June 2023 (UTC)
  • Support, especially on mobile. It's also useful when Google directly links to a specific section with highlighted text (with the `#:~:text=` location hash in the URL). a455bcd9 (Antoine) (talk) 12:01, 5 June 2023 (UTC)
  • Support. A. Parrot (talk) 14:14, 5 June 2023 (UTC)
  • Support It's about time that someone brought this up! Unlimitedlead (talk) 15:10, 5 June 2023 (UTC)
  • Support As long as it is discretionary. I would hate to have to reintroduce and re-link every person mentioned in a bio every section.--Wehwalt (talk) 15:33, 5 June 2023 (UTC)
  • Strong support; when writing lengthy articles, I've often felt more links would have aided the reader even if they had been reading top to bottom. Vanamonde (Talk) 18:32, 5 June 2023 (UTC)
  • Oppose For now, per Ian Rose. But establishing something in the MOS about more links on longer articles is something I could get behind. --TylerBurden (talk) 19:16, 5 June 2023 (UTC)
  • Support In lengthy articles, some things, people, and concepts may be mentioned in multiple contexts considerably removed from one another. Sinking of the Titanic is a good example—several individuals (e.g. J. Bruce Ismay, John Jacob Astor IV, and Benjamin Guggenheim) are mentioned early on and then again much later. Repeating the links at these later instances is helpful to readers, not harmful. It is especially helpful to readers who have skipped the earlier sections and gone straight to the sections they are interested in. Linking once a section is probably a bit excessive in many cases—obviously depending on section length—but it is a decent rule of thumb (my personal rule of thumb is one thousand words). TompaDompa (talk) 15:14, 6 June 2023 (UTC)
  • Conditional support if we also provide guidance on excessive linking. Linking is one of the newcomer tasks; tagging an article (or section) as requiring more links adds it to the list of articles for improvement on the Newcomers Homepage. The outcome can be dispiriting for those new editors, as they see their helpful attempts reverted as overlinking. If we state that repeat linking is acceptable because not everyone will read the article sections above, we should advise or warn there and then, in the same paragraph or even sentence because not everyone will read the MOS sections above, that it's not helpful to repeat every link at every opportunity and provide some guidance. We can't simply rely on common sense; that's learnt from advice and experience. If we do so guide/advise/warn, then this is a sensible, realistic and helpful proposal, even overdue, but if we don't then oppose. NebY (talk) 16:36, 6 June 2023 (UTC)
  • Strong support This is necessary in the era where the absolute majority of Wikipedia views are on mobile. Articles that follow current MOS linkage guidance are actively inaccessible to mobile readers; the MOS is concerned with accessibility and approachability in many areas, but falters on this. Even for desktop readers, it's not always trivial to figure out when an arcane concept was introduced (particularly for less technically competent readers who haven't figured out ctrl+f et al; keep in mind average tech literacy is shockingly low). This is compounded by long articles, articles on jargon-heavy subjects, long articles on jargon-heavy subjects, etc. I've supported this for quite a while and used it as practice already, MOS violation as it was -- but hopefully won't be for much longer. Vaticidalprophet 00:54, 7 June 2023 (UTC)
  • Support per above. Olivaw-Daneel (talk) 21:27, 7 June 2023 (UTC)
  • Support. In mobile and desktop, readers looking for specific information especially are faced with this issue. Having no wikilinks to important topics that were previously-linked can be a hindrance to the accessibility and convenience of how the information is presented. As with everything, it should be applied with common sense. I think it should note that this is generally best suited for long articles (a short, two-section article wouldn't need it), and that linking every previously-linked term in every section just because it can be isn't beneficial. The contextual value of linking something in any given section should be considered, which is often obvious. Lapadite (talk) 12:05, 13 June 2023 (UTC)

Discussion

Would this trickle over to MOS:SURNAME, where we would mention the full name of people in each section?—Bagumba (talk) 15:39, 6 June 2023 (UTC)

Personally, I wouldn't see that as a necessity: if I can click on the person's surname to find out who they were, I could do without their full name, and it might actually be a nice indicator that the person in question appeared already in the article. That's something I hadn't thought of when I opened this discussion though; it might be a good follow-up at that talk page if this passes. Folly Mox (talk) 15:57, 6 June 2023 (UTC)

Mobile and long articles I see two arguments for more re-linking: mobile and long articles. For phone view, all sections are collapsed (not so on my tablet). Someone jumping to a section, might miss an earlier link, thus the proposal to link once per section. However, this doesnt apply for subsections (e.g. 2.1, 3.1.2, etc), which are all viewable once its respective top-level section is uncollapsed. Does that mean this linking change generally applies to top-level sections and not subsections i.e. ink once per section, but not once every subsection. If so, this might also address some who had concern that there might be excessive linking if sections were (too) short.—Bagumba (talk) 16:49, 6 June 2023 (UTC)

The wording change I initially proposed was as smol as possible, to avoid proxility in the already formidable MOS, but I always had level 2 subheadings (the collapsey sort) in mind with the term "section". I suppose this could be explicated in prose or footnote. Anything deeplier nested than level 2 can't be collapsed, so appropriate links should always be identifiable by scrolling up and down. Tangentially, browsers tend not to search within collapsed sections, so even if using the browser's search function, you still have to manually expand all the sections for it to work properly. Folly Mox (talk) 21:49, 7 June 2023 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Clarifying the purpose of links

In a discussion on the Salmon chaos page, I was involved in a discussion with @Drmies because I did not understand the purpose of linking. They explained how things should work and I think the page should be edited so that the guidelines reflect these principles and prevent further confusion.

I interpreted this page to mean that pages should be linked if they are relevant (e.g. WP:BUILD's "Ask yourself, "How likely is it that the reader will also want to read that other article?"" or the lead's "Whenever writing or editing an article, consider [...] what links to include to help the reader find related information"), but they explained that pages should only be linked if they explain something. Obviously these are very different approaches. I have the immediate suggestions of:

1. Changing OVERLINK to read "We should not link to commonly used words and names, including the names of (current) countries" (etc).

2. Deleting "Ask yourself, "How likely is it that the reader will also want to read that other article?"" from WP:BUILD as this is irrelevant.

3. Deleting the first bullet point from UNDERLINK.

4. Generally, changing the guidance on principles and under- and over-linking to emphasize that we should link to specifics about which readers are unlikely to know, regardless of whether or not they help readers understand the page in general.

Does anyone have any further suggestions to help other editors? CohenTheBohemian (talk) 15:11, 21 March 2023 (UTC)

I object to the removal of the "unless particularly relevant" wording. Sometimes not linking an article just because you wouldn't link it elsewhere doesn't make sense. For example, because it's such a large and significant city, linking New York City is unneeded. But what if it's a topics specifically about the city, like the New York City Subway, or the Government of New York City, or one of the sports teams located there? Not linking the city in that case would be, frankly, dumb. So I'm reverting the changes for now, as no changes should be made while this discussion is just starting. oknazevad (talk) 16:36, 25 March 2023 (UTC)
Thank you for the reply, and letting discussion begin! (I was beginning to worry that nobody else cared...)
I originally felt the same as you, which is why I was confused at de-linking Taiwan from the Salmon chaos article, on the grounds that the place where something happened must be particularly relevant. But Drmies' advice was clear that this is how OL is supposed to work, and two admins, each having 15+ years editing experience, definitely outweighed my knowledge. So all my suggestions above are meant to make this clear rather than to change anything. CohenTheBohemian (talk) 15:17, 26 March 2023 (UTC)
Again, there is no list of places that should automatically be delinked, and whether or not a place is relevant depends on the subject of the article. Also, admins may be veteran editors, but they don't all agree on everything, and no one admin's interpretation is necessarily definitive. Which is why discussion is important. Especially when editing guidelines that have been hashed out over the decades. oknazevad (talk) 13:00, 4 April 2023 (UTC)
You are completely right, and I have done my best to discuss this - it is difficult to do without other editors' input. However, the guidelines clearly need clarifying (not changing) if they are as unclear as they seem to be. The alternative is bizarre.
Anyway, what do you think about changing the OL part to something like "should not be linked unless intrinsically relevant, e.g Government of New York City to New York City"? Hope you don't mind me borrowing your example - it's a good point!
CohenTheBohemian (talk) 12:57, 5 April 2023 (UTC)
I don't see a need to change MOS:OL, as it already contains the qualification that terms can still be linked if they're particularly relevant to the context in the article. As for the bigger question, it's a good idea to be parsimonious with links, but as the guidelines make clear, linking isn't only done for unfamiliar words. – Uanfala (talk) 13:20, 5 April 2023 (UTC)
Well, if they're not going to stand by their changes, I suppose that's that. No need for further discussion. Thanks to you and oknazvad. CohenTheBohemian (talk) 14:13, 8 April 2023 (UTC)
I concur with oknazevad's objection. I also object to removing "How likely is it that the reader will also want to read that other article". It is not irrelevant at all, but precisely the kind of judgment we use on a daily basis in determining what to link. Actually, I just oppose this entire revision proposal until we can see side by side (or, more probably, above and below) text to compare.  — SMcCandlish ☏ ¢ 😼  02:30, 14 June 2023 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Manual_of_Style/Linking/Archive_21&oldid=1171960380"