Misplaced Pages

Revision as of 00:28, 11 June 2014

This is the talk page for discussing improvements to the Manual of Style/Linking page.

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

Archives: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21Auto-archiving period: 2 months

Archives

WP:CONTEXT archives Support and opposition Archive 1 Archive 2 Archive 3 Archive 4 – repeating links Archive 5 – links in quotations Archive 6 Archive 7 – final archive WP:BUILD archive Archive WP:MOSLINK archives Archive 1 Jul 2006 and earlier Archive 2 Archive 3 Sep 2006 – Jan 2009 Archive 4 Jan 2009 – May 2009 Archive 5 May 2009 – Jun 2009 Archive 6 Jul 2009 – Aug 2009 Archive 7 Sep 2009 – Feb 2010 Archive 8 Feb 2010 – Jun 2010 Archive 9 Apr 2010 – Sep 2010 Archive 10 Sep 2010 – Nov 2010 Archive 11 Nov 2010 – May 2011 Archive 12 May 2011 – Oct 2011 Archive 13 Oct 2011 – Apr 2012 Archive 14 May 2012 – Nov 2012 Archive 15 Nov 2012 – Sep 2013 Archive 16 Sep 2013 – present

This page has archives. Sections older than 60 days may be automatically archived by Lowercase sigmabot III when more than 4 sections are present.

What is "major" for the purposes of overlinking?

WP:OVERLINK says not to link "the names of major geographic features and locations".

This is a list from an existing article:

Florence, Italy, Australia, Benin, British West Indies, China, Costa Rica, England, France, Japan, Kenya, Mexico, Spain and Zambia.

It could easily be argued that some of the linked places are not "major", but where do you draw the line? Does it mean "likely to be able to be pointed to on a map by the average reader"? Doesn't that depend on the place of residence, level of education, and experience of the reader (something that varies widely)? Aren't we being asked to make a decision that could be taken to mean value or importance to the world?

This guideline results in a significant number of edits and disagreement among editors. Is this really worth doing, just to avoid a few extra links in an article?

If I believed in this rule, I could easily unlink all of these as being "major", yet I would link Saint-Quentin, Aisne, which I've never heard of (but is a fairly large commune), which might seem stupid to many Frenchmen, who may feel quite justified when they come along and unlink it as being "major". —— 03:58, 26 March 2014 (UTC)

Good question. I would also approach the matter by asking, "how relevant is the content of that wikilink to the actual article?" Perhaps it could be put some other way instead of using a wikilink? For example, in biographical articles, a person might indeed have lived in - let's say - Saint-Quentin, Aisne at some point of his/her life, but how relevant is the content of ] as a whole to the article? Could it be possible to say, e.g. "...Saint-Quentin, a commune locating in the Aisne Department in Northern France; ..."? We are on a very slippery slope if we only think about about one's assumed level of knowledge.

I'm not sure if this is helpful but in my last thread on this Talk Page, In-text wikilinks possessing a serious threat, I was comparing the use of wikilinks with the use of in-text references in a paper-published articles: would you really add an in-text reference to an external source after every single geographical location (town, city, region etc.)? Wouldn't it appear somewhat ludicrous?

In my humble opinion, I concluded that when researching a paper-published article, indeed one wouldn't be able to do that. On the contrary, the author should phrase and write the text in such a manner that each piece of special terminology that is being used, is also being described to the reader. So forth, no disfavors would be made and the author would stay alarmed to keep the text easily-readable to the readers. What holds with the paper articles, also holds with wikiarticles.

Aside from an individual article, there are also few points that we should notice considering Misplaced Pages as a whole. Excess linking leads to:

1. quadratic explosion in the total number of links
2. decrease in the quality of articles making them hard and unpleasant to read

Trying not to talk too much, I just add one more citation from what I wrote earlier. I hope it'd demonstrate my point even just a little =P It goes as follows:

"I personally think it like this: if you are writing an article on paper, how many times and in which occasions you would like to add brackets referring to an external source? E.g. Volkswagen is a German (see Bayer, H.: World Geographic Atlas, 2013) automobile manufacturer headquartered in Wolfsburg, Lower Saxony, Germany (see Dickinson, R. E.: Germany: A General and Regional Geography, 2011). Volkswagen is the original and top-selling marque (see Kotler, P.: Principles of Marketing, 2006) of the Volkswagen Group, the biggest German automaker and the second largest automaker in the world.

I am not sure if the above-mentioned serves as the best possible example, but I think it demonstrates quite well the situation with excess linking. If one sentence has even up to three wikilinks (and many times even more), and this keeps recurring from sentence to sentence, the text would be absolutely ludicrous! In a paper-published article you couldn't naturally do that. What I think, is that the same that applies to publishing a paper article, applies to Misplaced Pages's articles too."

Anyway, thanks for your post AlanM1, you are bringing forth a serious topic here! Jayaguru-Shishya (talk) 10:11, 26 March 2014 (UTC)

The patchy effect is regrettable, although inescapable in some contexts (usually not as bad as that). Question: are links to any of those countries helpful in the context? It would be fine to link all if they are, I believe. Yes, there's a grey area where we end up having to make judgements as to how much our readers—first- and second-language speakers—are familiar with. Tony (talk) 11:08, 26 March 2014 (UTC)

If the links were rendered as in-line parenthetical clutter or even footnotes, I might agree that they are distracting and should be more limited, but when they are simply a slightly different color, ...? I quite often have reason to navigate to articles (especially those on places) that are just tangentially related to the subject, and am annoyed when those links are not there, especially when they were originally there and then removed by someone, forcing me to open a new tab, nav to wikipedia, type the name in the search box, and then select it from the results. —— 03:38, 17 April 2014 (UTC)

They're not just "a slightly different color" from the black default text: they're a garish blue—a hue that Tim Starling told me at a WM dinner years ago could well have been chosen by one of the early technical people on a whim. I'd have chosen a blue two shades more like the black (see my user page for how to alter the colour), with an option to make it the current garish blue in prefs, if your vision really wants the disruptive effect. Better still would have been plain black that underlines and is clickable when you hover over it—and every single word linked automatically. This is the case for my Encarta desktop dictionary: you get zero visual obstruction and maximum linking, all in one. The disadvantage, and it is a big one, is that it removes the current service we give readers in isolating our suggestions for the most valuable links. I'm afraid that one-click links to targets that "are just tangentially related to the subject" would litter the text, under the current system, with blotch-blue. Look at many articles in foreign-language WPs for that—the French, the Italian, the Russian, for starters; they seem to have no serious system for using wikilinking intelligently (i.e. selectively), even though most have some form of guidelines, usually as vague as water. Editors on those sites who argue for more disciplined wikilinking are usually howled down: it's a great pity. Tony (talk) 06:12, 17 April 2014 (UTC)

I agree with you Tony (talk) . In Finnish Misplaced Pages: the more links, the better. For example, if there are dates October 5th and 2004, they are both linked as October 5th, 2004. There is also actually an article for breakfast, where words, such as bread, porridge, fruit, berries, yogurt, cheese, and eggs are all linked :) I'm not kidding!

AlanM1, how does linking a geographic location help you to gain a better understanding on the main article anyway? Okay, you don't know for example where Nicaragua or Saint-Quentin, Aisne are located. Then you click the link and some 9000 words will pop-up in front of your eyes telling you everything about the natonal/regional economy, administration, population, transport, politics etc. (call me a no-life, but I just copy-pasted the whole deal into a words-calculator). Now, how relevant can you consider that wikilink? How much does that linked article itself help the reader to understand the main article he was reading?

Also, would you add to a printed article an external reference after such location? If we truly want Misplaced Pages to compare with distinguished encyclopedias (where we have some prominent results, so far), the exactly same applies with the conduct of writing wikiarticles. Please also bear in mind that there will be a quadratic explosion in the number of wikilinks if we indulge in excess linking. Jayaguru-Shishya (talk) 07:30, 17 April 2014 (UTC)

Cities, states, provinces?

I was about to ask the same question, but my question has to do with cities and states. Some editors consider all states as "major locations", and some consider some large metropolitan cities (New York, Los Angeles) as "major". I don't like when an MOS is so vague. --Musdan77 (talk) 19:29, 12 April 2014 (UTC)

My problem is that I don't know how we can justifiably call "all U.S. states" major but not all English regions/counties. What about Indian states? Nigerian states? A large number of people would consider each of these 1st-level admin divisions as "major" and unworthy of links, yet most of the world may not have heard of them:

• New Hampshire and Rhode Island
• Cumbria and Hertfordshire
• Jharkhand and Odisha
• Kwara State and Bauchi State

—— 03:38, 17 April 2014 (UTC)

As a "rule of thumb" if having a page linking to the state or other top-level entity would help the reader of the article in which the location is mentioned, then linking is appropriate. If it's only tangentially related to the topic and your average reader will at least have heard of the place, then no linking is required. I would link the 8 examples above in almost all articles, but I would not necessarily link London in an article about a person who happened to be born in London, England unless there were details in the city's article not known to the average reader which would help in understanding the person whose biographical article I was editing. Vague? Yes. When in doubt, I generally link such places in their first usage. davidwr/(talk)/(contribs) 21:22, 19 April 2014 (UTC)

What generally should not be linked?

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

I'd like to open discussion about the last addition with boldings. I don't think it is necessary to repeat a link after the lede. For example in many movie or band articles, there are usually members and castings listed in the lede. Aside from that, they are usually mentioned in various infoboxes also. If we now stick to the practice of linking everything that has already been linked in the article lede, we will end up with 3x the same link in total.

Most of all, we should always remember that every time we add a wikilink in the article, we send the reader a message: "There is something really important you should read about behind this link". However, in most cases the actual linked content is just tangentially related to the subject. Now, to link even once needs to have good reasons behind it; it seems highly unlikely that we would have good grounds to make such a claim to the reader thrice already.

Therefore, I'd like to suggest that the aforementioned passage is changed to:

''Generally, a link should appear only once in an article, but if helpful for readers, links may be repeated in infoboxes, tables, image captions, and footnotes." Jayaguru-Shishya (talk) 07:48, 22 April 2014 (UTC)

I strongly disagree with that proposal. The important part is if helpful for readers, and it's a good point. There are many times, mostly in lengthy articles, when it's convenient to link to a complex subject a second time. Sometimes a topic is mentioned in the lead, and then not again until 100,000 bytes later; if the subject is of significance to an article, I would prefer the option of occasionally creating a second link. As for the implied importance of a wikilink, we should not assume that readers are incapable of managing their own interests. Wikilinks provide more info for readers at their own option. We are not sending the message that they should read about it, merely that they can read about it if they so choose. Grayfell (talk) 07:08, 24 April 2014 (UTC)

Thanks for your response Grayfell! I may disagree with you, but I strongly appreciate your opinion! :D Jayaguru-Shishya (talk) 18:35, 24 April 2014 (UTC)

Likewise. I appreciate the sentiment, and agree that overlinking is detrimental to the project. I think the they way to address it is to follow the current standard, rather than changing it to be more limiting. Grayfell (talk) 20:26, 24 April 2014 (UTC)

For lengthy articles 3x links are appropriate. The word vital is wikilinked in another section of the article far way from the history section. The word "vital" is also unfamiliar to most readers. See WP:UNDERLINK: articles explaining words of technical terms, jargon or slang expressions/phrases—but you could also provide a concise definition instead of or in addition to a link. If there is no appropriate Misplaced Pages article, an interwiki link to Wiktionary could be used. User:Jayaguru-Shishya, I don't understand the reason why the wikilink was removed when you consider it fixes a technical word that is underlinked. QuackGuru (talk) 20:21, 26 April 2014 (UTC)

That sounds pretty much the wrong way around. I'd say only link if relevant. If explanation is needed, then make it plain on the face of the article or gloss to wiktionary for comprehension. Avoid links that detract from encyclopedic understanding of the subject and dilute the effectiveness of the linking mechanism. -- Ohc  20:41, 26 April 2014 (UTC)

QuackGuru, this is not the right forum to discuss edits of some certain article. Please take discussion to the place where it belongs to. Jayaguru-Shishya (talk) 15:25, 8 May 2014 (UTC)

QuackGuru: we expect readers of the English Misplaced Pages to know what "vital" means. Convenient desktop dictionaries are now free for all computers, too. Tony (talk) 15:55, 8 May 2014 (UTC)

Links on/off button?

I don't know whether this is technically possible. If it is, I don't know if it's politically possible. But it's worth at least throwing out there and getting shot down, since it would render most of this discussion moot.

What if there were a button to toggle the links off and on?

The page would be rendered initially as it is today. If the links were hurting my readability, I could choose to turn them off.

Whether that would mean making them non-links, or simply changing the color of the text, would be a purely technical consideration, and I think the latter would be better if it's possible.

What if I then wanted more information about a word or phrase in the text? That would depend on the implementation. If only the color were changed, then the links would still be there and functional. I could hover over the word/phrase; if it became underscored, and my mouse pointer changed to the "hand", and the tooltip popped up, I would know it was a link and I could click on it as usual. If, on the other hand, the implementation was to change it to a non-link, then it would be necessary to toggle the links back on.

It might or might not be necessary to re-render the page when the button is pressed, I guess that depends on the capabilities of HTML. But I wouldn't see that as a serious problem, since the majority of pages render fairly quickly. Mandruss (talk) 09:37, 12 May 2014 (UTC)

You can apply a local stylesheet (in Special:Mypage/skin.css, or locally, in your browser) to style links however it suits you. You could also use a (browser-speciifc) plugin to disable all links, on a page-by-page basis. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:06, 12 May 2014 (UTC)

Thanks, Andy. I guess I shouldn't have used the first person personal pronouns. This isn't about me and my readability issues, it's about a general solution to the problem of wikilinks and readability. Mandruss (talk) 11:13, 12 May 2014 (UTC)

You can certainly tone down that default garish blue: see my userpage ... scroll down. Tony (talk) 11:49, 12 May 2014 (UTC)

Allow me to suggest that we limit ourselves to solutions that will work for all Misplaced Pages readers. First, they need to be aware that a solution is available. Then they need to be capable of using it. Mandruss (talk) 12:00, 12 May 2014 (UTC)

The solution to overlinking is to remove excessive links, if there are any; but my response was based on the fact that what you (or any individual editor) prefer to see is not what others do. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:03, 12 May 2014 (UTC)

Agree on your first statement, but let's consider the case where there's no significant overlinking, just a lot of appropriate links. It happens a lot, and it's what at least some of this page is about. It's what this solution addresses. I hope you meant reader rather than editor; I don't think anyone discussing the issue of readability is talking only about editors. And regarding your last statement, I think this solution allows everyone to have it their way --- as well as change "their way" with the click of a button. Mandruss (talk) 13:19, 12 May 2014 (UTC)

Some interesting issues are intersecting here. Let's think about some extremes:

1. We could go the way of my desktop Encarta dictionary (which incidentally has a button taking you to the WP article of the same title you've typed into the dictionary/thesaurus); in Encarta every single word functions as a link, but professional appearance and smooth reading are not degraded by patchy blue spattered all over the place—you only see a word turn (light) blue and underlined when you hover your mouse over it; until then, it's plain-sailing black. What Encarta sacrifices—if it was appropriate (which it's not)—is the opportunity to display intelligently selected and rationed links to readers. It's not an idea that appeals to me in relation to our environment. This selecting is what wikilinking does so well on WP, when not used to excess and when skilfully done.
2. We could abolish linking. No one is thinking of that, I hope.
3. We could go back to the old days of linking anything in sight, as many (most?) other language WPs still do, today. I think the en.WP community has spoken on that one, now that people have become used over the past eight years or so of gradual taming of the linking system. I get a lot of echo-thank-yous, from editors I've never heard of, for gnoming out overlinking nowadays; it surprises me.
4. We could give readers on/off buttons, yes; that proposal would need careful thought, and I'm too attached to the maintenance of a good linking system to think this is a viable option. If linking is carefully applied and not too disruptive of the reading process, perhaps that wouldn't be necessary. See No. 5 ...
5. We could float the idea of a button or two that offer toned-down link hues. Perhaps: "Light-blue link color ◌ "

   I've become so used to the colour patch/thingy I installed about six years ago that when I go to other WMF sites, links are like neon signs. See possible gradations here. But I'm such a tech dummy I can't work out how to update the advice at the bottom for vector (it's still says "monobook"). BTW, many people wouldn't be happy with my choice of minimal difference between the link colour and the default black text, but there are gradations slightly sharper gradations that are a little closer to the current default. I'm interested in people's opinions generally about pref. buttons for link colour—and link colour in general. Apparently the colour was chosen without much thought by a WMF engineer one day back early this century (according to Tim Starling, when we spoke a long time ago about it). Tony (talk) 15:02, 12 May 2014 (UTC)

I wouldn't have a problem with your option 5, with just one little change: the option to go all the way black. Is there a reason I'm missing why that would be a bad thing? Mandruss (talk) 16:00, 12 May 2014 (UTC)

Greetings! I am glad to see that the issue is being discussed here. I have to say, however, that I agree with Tony1 on this one. It's more about displaying intelligently selected and rationed links to readers. In one word, relevance. Moreover, in most cases the actual linked content is just tangentially related to the subject, so there isn't really any reason to link such content. In my humble opinion, the main issue here is good linking skills. Jayaguru-Shishya (talk) 16:17, 12 May 2014 (UTC)

I think that convenience (to the reader) ranks with relevance as the essential criteria as to whether some link should be added. However, that wasn't the initial topic of this section; should we open a new section for that? ~ J. Johnson (JJ) (talk) 19:59, 12 May 2014 (UTC)

Links are for convenience. But the selectivity (based on editorial skills/judgment) is a large part of their utility for readers. Just maximising "convenience" would result in every single word being linked, and then you lose the selectivity benefit. Comment on Jayaguru's post: not relevance alone, but utility, which is a tougher test. Part of utility is the obviousness of link pipes (misleading pipes are a major problem I sometimes have to fix when gnoming); and specificity: fixing vague pipes gets to be too much for gnoming, since often you have to research sections and offspring articles to identify a more specific target). So quite a few additional criteria are involved in the balance sheet of whether to link, and if so, how. It's a skill like prose-writing that wikis have made possible ... indeed I would say they demand it for high quality. Tony (talk) 03:42, 13 May 2014 (UTC)

By convenience I was thinking in terms of how many places a link should be replicated such that a reader doesn't have to scroll back several lengthy sections to find the link. A link should not be inserted anywhere if it is not relevant, and even a relevant term should not be linked on every use. The interpretation of "only once per article" is being used to override editors' considerations of other criteria, wherefore I think we should explicitly allow "once per section". But we have wandered off-topic; perhaps this should be discussed elsewhere? ~ J. Johnson (JJ) (talk) 23:04, 14 May 2014 (UTC)

Let's do discuss that though; I concur that "once per page" is too rigid a rule, though "once per section" is too loose; typically once per page is enough, but once per page as a default is fine, if we permit re-linking in subsequent sections if it seems particularly pertinent to do so, especially in longer articles.  — SMcCandlish ☺ ☏ ¢ ≽ⱷ҅_ᴥⱷ≼  02:13, 16 May 2014 (UTC)

I think a clear guideline is necessary since otherwise there would be a link on every mention. I appreciate your opinions but personally, I am in favour of once in an article. Jayaguru-Shishya (talk) 13:29, 16 May 2014 (UTC)

Okay, I am going to work up a proposal; will bring it up in a new section in a day or two. ~ J. Johnson (JJ) (talk) 21:16, 16 May 2014 (UTC)

• But my default assumption is that readers start at the opening and read through. If they want to poke around patchily, they should expect lots of the same item to be blued out: they should be assume to instinctively know that skimming back is requred—or they can type into the search box what they want. The slippery slope would return us to the link-anything days, whereas it's been a lot of work to get the system pared back to where each link really means something. Tony (talk) 03:43, 17 May 2014 (UTC)

Let's continue this discussion below at #Proposal to allow links once per section. ~ J. Johnson (JJ) (talk) 20:37, 18 May 2014 (UTC)

The iceberg junk code

This recent addition strikes me as something that needs to be removed, and certainly does not reflect WP consensus or normal practice:

• Items within quotations should not generally be linked; instead, consider placing the relevant links in the surrounding text or in the "See also" section of the article. Alternatively, immediately after the quotation and its references, consider adding an {{Efn}} template for each term you're linking and explain its meaning (usually by getting it from the linked-to article's first sentence); and before the References or similar section add a Notes section with a {{Notelist}} template. Example:

Classical arts reviewer Chris Ng wrote, "the mural shows icebergs with great sensitivity."{{Efn|], a large piece of ice that floats freely after it broke from a glacier or ice shelf}}

== Notes ==

a. ], a large piece of ice that floats freely after it broke from a glacier or ice shelf

I don't think I've ever seen such an absurdity in any article here, and this kind of abuse of complicated footnote code "bloat" is far worse than any excessive bluelinking problem.  — SMcCandlish ☺ ☏ ¢ ≽ⱷ҅_ᴥⱷ≼  00:47, 14 May 2014 (UTC)

Oh my. Yes, that is awful. Let's delete it (or at least decently bury it) before it propagates. ~ J. Johnson (JJ) (talk) 22:45, 14 May 2014 (UTC)

I'm going to WP:BOLDly just delete that nonsense.  — SMcCandlish ☺ ☏ ¢ ≽ⱷ҅_ᴥⱷ≼  02:13, 16 May 2014 (UTC)

Proposal to allow links once per section

This proposal arises from a discussion that developed above at #Links on/off button.3F, and a discussion at WT:Manual of Style/Layout#See Also section: is our guidance realistic.3F that came down to a question of overlinking.

At WP:OVERLINK the MOS says: "Generally, a link should appear only once in an article..." (emphasis in the original). Some editors have ignored the "generally" and interpret this strictly, using this as grounds for removing all "duplicate" links, including those in a separate "See also" section. Many editors find the strict "only once" interpretation too rigid, while others are concerned about (to quote Tony) a "slippery slope would return us to the link-anything days". A reasonable compromise is to limit links to only once per section. Therefore I propose that the last two sentences at WP:OVERLINK be replaced with the following, which slightly relaxes the rigid interpretation, and clarifies some related aspects:

In general links should be made only at the first instance in the text of an article or section (not counting instances in infoboxes, tables, image captions, or footnotes). Where a link is proper at different places in an article it may be repeated as a convenience so the reader does not have to scroll back through the article to find it. Links that are otherwise acceptable should not be removed solely because they are duplicated, but only if duplicated excessively, such as multiple times in a given section.

"See also" sections are intended as a convenient guide to articles on related topics that a reader might be interested in, independently of whether the such articles have been mentioned or linked in the current article. Links are not precluded from a "See also" section because they have appeared earlier in an article.

~ J. Johnson (JJ) (talk) 20:47, 18 May 2014 (UTC)

Support/Oppose

• Support as proposer. ~ J. Johnson (JJ) (talk) 17:36, 24 May 2014 (UTC)
• Oppose—I think this is very unwise. Tony (talk) 03:14, 19 May 2014 (UTC)
• Oppose It is specifically discouraged to have a link in the 'See also' section that is already in the article itself. -- Ohc  03:31, 19 May 2014 (UTC)
• Oppose as a terrible idea. Specifically ruled out by WP:SEEALSO. --John (talk) 10:00, 19 May 2014 (UTC)
• Support the first paragraph, oppose the second. As a reader I really hate it if I have to go link hunting, i.e. if I have to manually scan the article text to find a link just because the term was already linked somewhere else. E.g. an article about a composer might mention specific works at different places, and then it would be really inconvenient if only the first place were linked. Tobias Bergemann (talk) 10:17, 19 May 2014 (UTC)
• Oppose This would take Misplaced Pages back to the Stone Age of linking. Even the current policy is quite a flexible one. Jayaguru-Shishya (talk) 16:30, 19 May 2014 (UTC)
• Oppose - if people are ignoring part of what's currently there, you should discuss the matter with them. The current wording is flexible enough to allow relinking where it makes sense to do so; the proposed wording is far too loose. Nikkimaria (talk) 20:45, 19 May 2014 (UTC)
• Support the first paragraph for the reason given above by Tobias Bergemann. Many readers jump to a section and will not have had an opportunity to benefit from an earlier link. (So far all the "oppose" arguments seem to be "we've always done it that way." Not particularly persuasive.) Butwhatdoiknow (talk) 21:34, 19 May 2014 (UTC)
• Oppose - I didn't think Misplaced Pages was a democracy! 64.134.232.217 (talk) 18:34, 24 May 2014 (UTC)
  • It isn't; which is why you are supposed to give your reason, not merely "vote". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:47, 27 May 2014 (UTC)
• Support - the policy of linking once only is a nonsense, particularly in lengthy articles - linking within sections makes much more sense. Readers will often jump to a specific section (which is what the TOC is designed for) so why force them to go hunting through each other section to find a previous occurrence of the word(s) to find a link to the relevant article? The mobile apps of WP present an article with all the sections closed up and actively encourage section jumping and so exacerbate the problem. It is also an extremely poor experience for users of screen readers who will find it very laborious to use link nav to scroll up and down the page. Bladeboy1889 (talk) 10:26, 26 May 2014 (UTC)
• Support first part - much more helpful to the reader. Consider also mobile view, where sections are collapsed, and a link in, say, section 1 may not be seen by someone who has expanded, say,only section 4. Oppose second part. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:47, 27 May 2014 (UTC)
• Oppose The issue seems to be the implementation of the current policy, rather than its actual wording, and I am nervous about how the altered wording would be applied in practice. Hchc2009 (talk) 14:36, 27 May 2014 (UTC)
• Oppose Is the proposal a solution to a nonproblem? 166.147.88.28 (talk) 02:47, 6 June 2014 (UTC)
• Support This change would address two annoyances I have frequently seen in longer articles. The purpose of links is to help a reader quickly finding associated information without having to guess the exact article title and enter it into the search box. Complex articles aren't read linearly. Users might skip chapters by clicking the TOC, or they might arrive at the article via a link from another article or redirect to a section header or embedded anchor. A strict rule for no repeated links defeats the very purpose of why links exist in the first place. From the usability point of view, a reader cannot be expected to hunt previous chapters for links - a reader might waste alot of time scanning through information outside his/her scope of interest trying to find a link without even knowing if it exists. This is a very bad user experience. Therefore, the question, if a link should exist or not, should be answered from the viewpoint of a reader arriving at a section header without knowing the previous part of the article. In a very short section or article, it might still be okay to scan the whole article for a link, while in longer and more complex articles it is not. Similar for See also sections. While it is generally unnecessary to repeat already existing links in the See also section, sometimes See also sections contain well-organized lists of orthogonally related topics and if some of the links, which would belong into such a list, cannot be found there (because they are already used further up), it could easily be interpreted as if the subject isn't covered at all in Misplaced Pages. In such cases, not repeating the link creates inconsistencies and hinders easy navigation. --Matthiaspaul (talk) 20:45, 8 June 2014 (UTC)
• Oppose It opens the door to large scale overlinking and repetition of large number of links in the see also section that are already in the text.--SabreBD (talk) 23:14, 8 June 2014 (UTC)

You didn't provide a rationale for your comment. My request for further comment was moved to the bottom of the page; perhaps you would comment there. ~ J. Johnson (JJ) (talk) 00:14, 11 June 2014 (UTC)

Discussion

Invoking the "not repeat" language at WP:SEEALSO is meaningless as it only repeats the "only once" langauge here (from which it is derived). It most certainly does not rule out making changes. Nor does it rule out "duplicate links": the supposed prohibition is given (in both places) as a general rule. The problem is that quite a few editors routinely remove such "duplicate" links, without consultation, as being flatout prohibited by WP:OVERLINK (end of discussion, editorial judgment and common sense be damned). The current language is nominally general and flexible, but the result is not.

You opposers seem to fear some immediate and overreaching kudzu-like fate for en:Misplaced Pages if ever a link be repeated in an article. Note that this proposal in no way relaxes the requirement for relevancy, nor what should not be linked. There has been no showing how duplicating a link in "See also" is in any way objectionable, except for violating an arbitrary and rigid prohibition. If this is such a terrible and unwise idea please show how that is so. ~ J. Johnson (JJ) (talk) 00:18, 20 May 2014 (UTC)

We don't "fear", we know. Less is more. Links are valuable when properly used; they detract when in excess. Although overlinking has reached its peak, it would be a shame to let those gains in parseability/readability be eroded by removing provisions in the guideline that have served the project well. In the same way as WP:MOSFLAG allows the use of flag icons in clearly defined but limited circumstances, editors have naturally gone beyond what is allowed or optimal, and linking practices are no exception. Links far enough away from the first one are not prohibited as it seems the nom implies. Thus the rationale for allowing an existing link to be repeated in the See also section is mighty weak, and is likely to turn it into a repository of repeat links. -- Ohc  01:24, 20 May 2014 (UTC)

Unfortunately, this started out as two proposals: One to make it clear that links may be repeated in articles. Another to say that links can be repeated in See also. Are you two discussing both proposals or only one? And if one, which one? Butwhatdoiknow (talk) 12:29, 20 May 2014 (UTC)

  No, it is really only one proposal. The second paragraph clarifies a special case. The existing OVERLINK text says once per article, which implies the whole hog, and then exempts infoboxes, tables, etc., without addressing "See also" sections. The proposal specifies the text, which would not cover a SA, but Wikipedians being so various in their interpretations I thought it best to explicit address SAs. I could have proposed merely adding "See also" to the list of exemptions, but it seemed to me better to go for the most general case.

  Ohc: Setting an arbitrary limit ("only once"!) on the number of links without regard to their quality or relevance has a serious failing: it gives a free pass to link ANYTHING (with a few exceptions) once. Provide any example you want, and identify every instance of what you deem overlinking, and I suspect that many of those links (most?) will be 1) non-duplicated links, and therefore pass under the existing rule, and 2) better handled (i.e., removed) on the basis of relevance and quality.

  As to turning a SA "into a repository of repeat links": do you mean like repeating every link in an article? The way to avoid that is not by absolutely and universally prohibiting repetition, but addressing that problem at WP:SEEALSO. ~ J. Johnson (JJ) (talk) 22:57, 20 May 2014 (UTC)

"It gives a free pass to link ANYTHING"? No, not really. It's been described at sections What generally should be linked and What generally should not be linked. In my humble opinion, the latter should also be made more precise even: it's still missing many problematic areas that should be added on the not to be linked list. For example, I hope WP:OVERLINK could be extended to geographic locations in general (towns, cities, regions, countries). A small quote from my post at What is "major" for the purposes of overlinking?:

... how does linking a geographic location help you to gain a better understanding on the main article anyway? Okay, you don't know for example where Nicaragua or Saint-Quentin, Aisne are located. Then you click the link and some 9000 words will pop-up in front of your eyes telling you everything about the natonal/regional economy, administration, population, transport, politics etc. (call me a no-life, but I just copy-pasted the whole deal into a words-calculator). Now, how relevant can you consider that wikilink? How much does that linked article itself help the reader to understand the main article he was reading?

.

Jayaguru-Shishya (talk) 00:18, 21 May 2014 (UTC)

• What u talking about, Dude? WP:UNDERLINK says

  An article is said to be underlinked if words are not linked that are needed to aid understanding of the article. In general, links should be created to:

  *relevant connections to the subject of another article that will help readers understand the article more fully (see the example below). This can include people, events, and topics that already have an article or that clearly deserve one, so long as the link is relevant to the article in question. (emphasis is mine)

  That links need to be relevant and useful to the reader's understanding is unambiguous, and the converse being that anything that isn't strictly relevant and useful to the understanding of the subject would constitute overlinking. -- Ohc  03:39, 21 May 2014 (UTC)

That's exactly my position: that relevancy (or usefulness to understanding) is the proper criterion of whether a given link is in, not whether it's already been used. The current rule (as interpreted) is that no matter how strongly relevant, useful, etc., a link is, sorry, you can use it "only once" (emphasis in the original) in the entire article. And if editors won't assess links on qualitative relevance and usefulness then, yes, any link can get a free pass once. ~ J. Johnson (JJ) (talk) 21:05, 21 May 2014 (UTC)

• And if editors won't assess links on qualitative relevance and usefulness then, yes, any link can get a free pass once – Your fears are overblown and based on a false hypothesis. I see plenty of anecdotal evidence, on a daily basis, of editors applying the "relevancy test". -- Ohc  00:19, 24 May 2014 (UTC)

What are you talking about? If you're saying that editors do assess links on the basis of relevancy, then fine, we don't need a fixed, quantitative limit of "only once". On the other hand, it is a fact that "only once" is being applied by some editors strictly and without regard to relevancy (resulting in non-hypothetical aggravation). I'm saying the proper criterion for assessing links is, as you said, "relevant and useful to the reader's understanding". At best "only once" is just a crutch to avoid having to look at relevancy. ~ J. Johnson (JJ) (talk) 18:14, 24 May 2014 (UTC)

Relevancy deals with the quality of the link. About the quantity, it is said that:

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

I don't see any reason to loosen the current one. There are editors out there who would like to link a specific term on almost every single mention, and I don't find it a good idea to encourage that via WP guidelines. Jayaguru-Shishya (talk) 13:57, 25 May 2014 (UTC)

  JS: Not eactly. What the page says is "... only once" (you left off the emphasis). And I have shown you the reason for changing this: some editors intepret this as absolutely only once, regardless of quality. Quantity, to the measure of only once, trumps quality, relevance, helpfulness, etc.

  Your objection to linking "a specific term on almost every single mention" is ill-formulated. If a term is mentioned only once (and assuming a link is relevant and useful), would you object because every instance (albeit only one) is linked? What is the precise problem you envision with linking "on almost every single mention"? ~ J. Johnson (JJ) (talk) 00:53, 26 May 2014 (UTC)

• As clearly stated in the guideline, a second appearance is not prohibited, but seems to be strongly implied by JJ; the assertion of only once is not borne out by any evidence – I have not seen editors removing a second occurrence outright when there is reason to keep same (for example, where an occurrence is in a section far removed from the first). OTOH, the proposal opens the door to linking of relevant words as many occurrences as there are sections in any given article. From my own observations about past linking practices, I do not consider this to be a desirable state of affairs. -- Ohc  01:42, 26 May 2014 (UTC)

"nly once" (with emphasis) is from WP:OVERLINK (second to last sentence), as just quoted by JS. Yes, it is supposedly qualified by the leading "generally", but in practice it is strictly interpreted by some editors as sufficient basis for removing all "duplicates". Evidence? see (e.g.) this edit and this edit for "outright removal" on the sole basis of being duplicates. ~ J. Johnson (JJ) (talk) 21:31, 26 May 2014 (UTC)

Ohconfucius: Do our different views here resolve (in part) to your "I have not seen editors removing a second occurrence ...", whereas I have seen them? You asserted that "the assertion of only once is not borne out by any evidence". I have provided two contra-examples where links have been removed on the sole basis of being duplicated; do you require more? ~ J. Johnson (JJ) (talk) 20:59, 4 June 2014 (UTC)

J. Johnson: You seem to have created this proposal as ex-post justification for edit-warring with another editor over their removal of overlinking. I have no problem with those removals from the See also section. I am only surprised that more links were not eliminated, but perhaps the editor was only working based on WP:SEEALSO. You just have to search around the rest of the article and count how many times the links are repeated after the 'See also' links had been removed to see what I mean. -- Ohc  08:00, 5 June 2014 (UTC)

You said you had "not seen editors removing a second occurrence", I supplied two instances, and on that basis you make a wild accusation of edit-warring? If that is the best response you can make then I think it is adequately demonstrated that "duplicate" links are in fact being removed on the sole basis of being duplicate. Moving on, I suggest that the question is now whether this is how WP:OVERLINK should be interpreted. ~ J. Johnson (JJ) (talk) 23:11, 6 June 2014 (UTC)

The examples you gave are not second instances, but nth instances. The Seattle Fault article isn't very long, "Tacoma Fault" is already linked once in the middle of the article, and once in the navigation template at the bottom, so the See also link is not warranted. User:Brianhe's edit was perfectly reasonable and also in line with guidelines. See this evidence of your edit warring to reinstate the link to Tacoma fault.

In the Puget Sound faults article, I found five links to Seattle Fault in the version you apparently object to, plus one instance using the {{main}} template. If these two examples are the best you can do to contest my assertion, I think we can close this discussion straight away. -- Ohc  02:51, 7 June 2014 (UTC)

Geez, if you wanted more/better examples you only had to ask. (The ones I previously supplied were just the easiest to find.) Fine, here are some examples that cite WP:OVERLINK as authority for "only once" where I have no involvement:

• Edit summary at Plate tectonics: "Alfred Wegener's name is already linked twice in the article - so for strict compliance, one of those links should be removed, ...."

• Comment: "That was a duplicate link, removed per WP:OVERLINK".

• Comment: "Sorry, but every city is already linked, and that's why they're not supposed to be linked again. Per WP:OVERLINK, a link should really only appear once in an article unless the two links are a long way away from each other in the article."

• Comment: "According to WP:OVERLINK— A link should appear only once or twice in a page."

And there is a whole trove of "There are a couple of duplicate links which should be removed per WP:OVERLINK" comments from a script being used on GA articles, with subsequent removal. E.g.:

• Talk:Little Boy/GA1
• Talk:Battle of Caishi/GA1
• Talk:Fat Man/GA1

And many more instances of second instances being removed merely for being a second instance, as explicitly stated. Your assertion that "a second appearance is not prohibited" is contradicted by the evidence. ~ J. Johnson (JJ) (talk) 21:21, 7 June 2014 (UTC)

Okay, there are other examples, but are these examples of a serious problem? 173.160.49.206 (talk) 21:29, 7 June 2014 (UTC)

Well, Ohc has just accused me of edit warring (hardly a collegial attitude) simply because I and another editor have different interpretations of OVERLINK. Whether an absolute "only once" rule is helpful or not is frequently questioned, most recently here and here. ~ J. Johnson (JJ) (talk) 22:18, 7 June 2014 (UTC)

And you have not accused others of edit warring? 24.9.99.196 (talk) 23:38, 7 June 2014 (UTC)

An entirely irrelevant comment. With one exception (64.134.232.217) all of the IP addressees that have appeared in this discussion (166.147.88.28, 173.160.49.206, 24.9.99.196, and 166.147.88.48) have a common source (Denver, Colorado) and pattern of edits. If anyone else wants to discuss whether the "only once" interpretation results in a problem, fine. But I am disinclined to engage in an irrelevant discussion with an anonymous, wiki-hounding sockpuppet. ~ J. Johnson (JJ) (talk) 23:39, 8 June 2014 (UTC)

• @J. Johnson: Just to recap: I said: "I have not seen editors removing a second occurrence outright when there is reason to keep same (for example, where an occurrence is in a section far removed from the first)". I stand by that. Note that I'm not saying second occurrences need to be removed in an absolute sense. But when there's a third (or more) link to the same article from another, chances are at least one can be removed.
  1/ Plate_tectonics: As far as I can see, this does not involve link removal, and there seem to be an adequate number of links for navigation. Arguably, the link locations could have been swapped, but it's otherwise irrelevant.
  2/ User_talk:Sitush: The article referred to was Narendra Modi, and it seems that "The Emergency (India)" may have been a necessary and relevant link, once. A second instance, only three sentences later, was superfluous and correctly removed.
  3/ User_talk:Imzadi1979: That comment is completely in line with guideline and what I am to preserve. The inexperienced editor was placing links in the next line and here omitting the necessary disambiguation. In fact, he/she had been attempting to alter wikilinks to mine metadata for an external site.
  4/ User_talk:Moonriddengirl: Er, seems to illustrate a humour deficit on your part. Thanks for providing the comedy, amigo.

  You have not disproved my above (repeated) assertion. As for me, I don't have any more time for this nonsense. -- Ohc  04:17, 9 June 2014 (UTC)

  I can't answer for what you can, or cannot, see. The "assertion" I address is where (on the 26th) I said that "some editors interpret this as absolutely only once", and then you said "the assertion of only once is not borne out by any evidence." What I have shown is multiple assertions by various editors that WP:OVERLINK means "only once".

  You seem to have overlooked that this proposal would still prohibit duplication within a section, and would not prevent removal of links duplicated "only three sentences later", or "in the next line" (cases #2 and #3 above); there would be no lessening of authority to do so (a point I made previously). Editors would have to give up this "more than one" quantitative crutch, but there would still be adequate bases for blocking "large scale overlinking". ~ J. Johnson (JJ) (talk) 21:10, 9 June 2014 (UTC)

Discussion continues

John: how is this "a terrible idea"? ~ J. Johnson (JJ) (talk) 20:33, 23 May 2014 (UTC)

John: your bare opinion, lacking any explanation or reason, amounts to no more than WP:I just don't like it, and warrants no consideration. Can you provide any basis for your opinion? Or should should we ignore it? ~ J. Johnson (JJ) (talk) 21:03, 4 June 2014 (UTC)

Yes, and please yourself, respectively. --John (talk) 22:14, 4 June 2014 (UTC)

If you have some basis for your opinion that might inform the discussion I should like to hear it; perhaps it would even change my views. (And I allow that even feelings and speculative considerations can be relevant, albeit not compelling.) ~ J. Johnson (JJ) (talk) 22:32, 4 June 2014 (UTC)

Let's be polite. — Preceding unsigned comment added by 166.147.88.48 (talk) 02:27, 6 June 2014 (UTC)

The implementation issue dervies from the "absolutely only once" interpretation, which is based on the existing actual language of "only once" (emphasis in the orignal). ~ J. Johnson (JJ) (talk) 23:03, 6 June 2014 (UTC)

As explained in previous comments: there are problems trying to find a link when it is buried in the article's text. There are also problems when editors clash on whether the policy allows any exceptions to "only once". ~ J. Johnson (JJ) (talk) 23:03, 6 June 2014 (UTC)

Sabrebd, regarding your "oppose" (above):

Why do you think this "opens the door to large scale overlinking and repetition"? ~ J. Johnson (JJ) (talk) 21:14, 9 June 2014 (UTC) (added context to comment that JS moved. J. Johnson (JJ) (talk))

J. Johnson, you are misinterpreting the guideline. There is no such thing as only once restriction. If somebody still misinterprets the guideline, you can sure notice him/her about that per WP:MOS/Linking.

Anyway, I have already quoted the current guideline, and you have shared your own opinions about it. If there is nothing more to add to the discussion, I'd suggest to not go over the same arguments over and over again. Jayaguru-Shishya (talk) 22:22, 9 June 2014 (UTC)

I suspect we keep going "over the same arguments over and over again" because we are not really on the same argument, have yet to find a common view from which to proceed. Like, you say that I am misinterpreting the guideline. Only I am not interpreting it (mis- or otherwise). What I am saying is that other editors are interpreting the guideline (as shown by the edits and comments I have pointed to) as an absolute "only once" restriction. Whether this is a mis-interpretation is a separate issue; my point is that the guideline is, in fact and as demonstrated, applied as an "only once" restriction. If you don't see that (or can't explain why I should see it differently) then there is a problem. If you do agree that some editors mis-apply "only once", then we can proceed to the next sub-issue. ~ J. Johnson (JJ) (talk) 00:28, 11 June 2014 (UTC)