Wikipedia

Wikipedia talk:Manual of Style/Linking/Archive 13

< Wikipedia talk:Manual of Style | Linking
Latest comment: 12 years ago by Rich Farmbrough in topic Incoming links at India and Turkey
Archive 10Archive 11Archive 12Archive 13Archive 14Archive 15→Archive 20
Latest comment: 13 years ago17 comments5 people in discussion

I notice that many FA-quality articles repeat links twice: once at the first occurrence in the lead, and again at the first occurrence after the lead. I'm looking in WP:LINKS and I dont see mention of that. If it is a recommended practice, I would expect to see that rule stated either in the Lead section or the Repeated Links section of this page. Am I missing something? --Noleander (talk) 15:42, 29 October 2011 (UTC)

WP:REPEATLINK doesn't exactly apply, but I didn't find anything that comes closer: "exceptions ... where the later occurrence is a long way from the first". Art LaPella (talk) 22:36, 29 October 2011 (UTC)
Yes, I saw that exception. But I looked at the last three FA articles on the WP home page, and all three repeated links, even when they were very close, for instance in the first section immediately after the lead. For example in todays featured WP article, Lester Brain, the lead has links for New South Wales and Royal Australian Air Force, yet those links are repeated about ten sentences later, in the first section following the lead. I don't have an opinion for/against this "repeat the link at first occurrence after the lead" convention, but if the black-belts at FA are using it as a standard practice, I suggest that WP:LINKS mention that as a "best practice" or "commonly used in good articles" or something like that (but doesnt suggest that it is mandatory). --Noleander (talk) 22:45, 29 October 2011 (UTC)
(I guess I should mention where this came from: I recently submitted an article for GA review, and the reviewer asked me to repeat the lead links, and I complied, based on what I saw at FA. But I was surprised this convention was not mentioned in WP:LINKS). --Noleander (talk) 22:55, 29 October 2011 (UTC)
That's because WP:LINKS and the MOS generally mirror the views of a dogmatic clique of delinkers, and reflect neither good practice nor reality nor commonsense which says that repeat links are desirable. -- cheers, Michael C. Price talk 23:07, 29 October 2011 (UTC)
I take it, then, that you support changing LINKS to include mention of the convention found in many FA articles.  :-) Let me go ahead and make a formal proposal: I propose to amend WP:LINKS, in the WP:REPEATLINK section, to say something like, "it is acceptable to repeat links at the first occurrence after the lead". --Noleander (talk) 23:50, 29 October 2011 (UTC)
Sure, go ahead. -- cheers, Michael C. Price talk 06:27, 30 October 2011 (UTC)
Perhaps we could initiate an RfC to get input from additional editors? --Noleander (talk) 14:20, 30 October 2011 (UTC)
Latest comment: 12 years ago95 comments28 people in discussion
The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
Closing Rationale: The support argument has the numerical majority with 16 to 6. However more weighty in my reasoning is the fact that several of the opposers argue that the change would be an addition to the policy and as such perhaps a case of WP:INSTRUCTIONCREEP - I think this argument is compatible with the support argument - IF the proposed change is implemented in a way that gets rid of the rule and the list of exceptions, by substituting the use of commonsense in achieving a balance between overlinking and helpfully guiding readers towards articles of interest. In short the change should not stipulate any preferences for particular way of repeating links or not other than a general admonition to avoid overlinking, it should not introduce any new requirements or suggestions - only simplify existing ones. In my eyes this is therefore one of the rare chances we have to simplify a policy rather than complicate it - and I hope that when framed this way some of the opposers will look favorably on the closure as: "support simplifying rules for link repetition". Now I leave it up to you all to decide whether deleting the section altogether, merging it into another section (e.g. the one on Leads) or merely simplifying it is the preferred solution.·ʍaunus·snunɐw· 00:54, 26 November 2011 (UTC)

Many FA-quality articles repeat links twice: once at the first occurrence in the lead, and again at the first occurrence after the lead. For example in Lester Brain, the lead has links for New South Wales and Royal Australian Air Force, yet those links are repeated about ten sentences later, in the first section following the lead. More than half of recent FA articles follow this convention. Yet this is contrary to the advice in WP:LINKS, which suggests that only the first occurrence should be linked. Question: Should WP:LINKS be modified (probably in the WP:REPEATLINK section) to acknowledge this as a permissible (but not necessarily recommended) practice? --Noleander (talk) 14:28, 30 October 2011 (UTC)

  • Comment - I have no strong opinion one way or another. But something needs to change: either this MOS should acknowledge this as an acceptable convention (presenting it simply as an option, not a recommendation), or else the FA reviewers should be asked to change their habits, because it is contrary the MOS and confuses editors that are looking at FA articles as exemplars. --Noleander (talk) 14:29, 30 October 2011 (UTC)
  • Support - WP:REPEATLINK should allow editors the option of repeating links more often so that readers have the choice of using them.
    1. We shouldn't require articles to be read linearly.
    2. We shouldn't require readers to have perfect memories.
    3. We shouldn't require readers to know whether they might need/desire information later on in an article the first time they meet a link.
    4. We shouldn't think that we know the profile of readers; they're a varied bunch, in interests and knowledge; more links cater for this by providing more choice.
    5. Policy here should reflect commonsense and actual practice, not the anally-retentive views of a bunch of control freaks.
-- cheers, Michael C. Price talk 16:33, 30 October 2011 (UTC)
  • Support - The lead is meant to be a summarizing section and as such many who are interested in more in-depth knowledge will skip over it once they knew they are reading the right article. I have always assumed this is a WP:COMMONSENSE exception to not posting links more than once/article in the prose.∞陣内Jinnai 16:43, 30 October 2011 (UTC)
  • Comment I was always under the impression that it was acceptable to repeat links, in separate sections, as each section tends to stand on its own. Sometimes I am only interested in reading a particular section of an article, and I shouldn't have to scan back several sections in order to find a link.--JOJ Hutton 17:25, 30 October 2011 (UTC)
    • I think it also depended upon how close the terms linked were. If the previous section ended with a link and the next section started with that same link it would probably be considered too close together.∞陣内Jinnai 18:14, 30 October 2011 (UTC)
  • Sometimes. It is possible that someone reading a section of an article has not read the sections before it, so it makes sense to repeat some (not all) links. On the other hand, we can assume everyone reads the lead unless they've come to the article through a section link (or a section redirect). In any event, what the guideline says should match what the best current practice is; there's very little point in deprecating a practice if not even most featured articles avoid it. ― A. di M.​  19:41, 30 October 2011 (UTC)
    The RfC question is: Should WP:REPEATLINK be amended to acknowledge the convention seen in many FA articles? Which is a Yes/No question. So I take it that your "Sometimes" comment means "Yes, it should be amended, but the new wording should indicate that the repetitions are optional and should be considered in light of the article as a whole, etc". --Noleander (talk) 19:54, 30 October 2011 (UTC)
    Exactly. ― A. di M.​  22:18, 30 October 2011 (UTC)
  • Comment It seemed a bit strange to have this discussion without letting the FA people know. So I notified them. CambridgeBayWeather (talk) 21:18, 30 October 2011 (UTC)
Thanks. I had already notified them also at Wikipedia_talk:Featured_article_candidates#Question_about_repeat_links ... on the FAC page, which seems to have more activity. --Noleander (talk) 23:25, 30 October 2011 (UTC)
  • Support, many readers do not read articles linearly, but rather read the first sentence/paragraph, skip to the table of contents and go to their section of interest. In addition to the guideline allowing an additional link "where the later occurrence is a long way from the first" it should allow additional links when they are in different section from the first to accommodate the fragmented way our articles are often read.AerobicFox (talk) 01:56, 31 October 2011 (UTC)
  • Support there's a bot being written to check the lead and body separately for overlinking, so the bot will be working on that basis too. I've adopted this practice myself in my last few FAs Jimfbleak - talk to me? 06:42, 31 October 2011 (UTC)
  • Oppose. [Although opposing is irrelevant, since the current guidelines allow for this, using common sense. Tony (talk) 09:50, 31 October 2011 (UTC)] "For example in Lester Brain, the lead has links for New South Wales and Royal Australian Air Force, yet those links are repeated about ten sentences later, in the first section following the lead." This sounds like a really bad use of our wikilinking system. You mean, shove it in their faces if they didn't get 20 seconds beforehand? Readers are not fools. Tony (talk) 07:28, 31 October 2011 (UTC)
    • Wow, and you call that "shoving it in their faces"? -- cheers, Michael C. Price talk 07:37, 31 October 2011 (UTC)
      • This RfC is meaningless unless the wording is proposed. The difficulty is that the current "Repeated links" section doesn't specifically disallow the repetition of a link that first appears in the lead. It was deliberately cast that way for flexibility; but we rely on editors to use common sense. Just what is being proposed?Tony (talk)
        • And yet the inflexible and dogmatic implementation by Tony et al shows a lack of common sense (which is what people are complaining about), which is why such flexibility needs to be explicit in MOS. -- cheers, Michael C. Price talk 08:20, 31 October 2011 (UTC)
        • @Tony: I understand what you are saying, but when I read WP:LINKS, it says "In general, link only the first occurrence of an item" which, to me, means the FA practice is non-compliant. I agree with you that WP:LINKS should permit the kind of flexibility that we are seing in the FA articles. This RfC is simply asking that WP:LINKS wording be improved so editors, like myself, don't misunderstand it, and think the double-linking is prohibited. --Noleander (talk) 13:20, 31 October 2011 (UTC)
... also, Tony asked about what specific wording change is proposed. I deliberately left that out of the RfC description, because a particular wording might narrow the discussion down too far. One option is a new bullet in WP:REPEATLINKS, as in: "* When a link is in the lead, the first occurrence following the lead may also be linked." But that kind of detail can wait until if/when this RfC consensus turns out to be "Support". --Noleander (talk) 14:12, 31 October 2011 (UTC)
  • Support as a natural and appropriate use of the linking system, as editors can and will exercise appropriate discretion as they do with other aspects of the style guidelines. This should be left to the discretion of the editors of a page in accordance with what works best there. --Ckatzchatspy 07:47, 31 October 2011 (UTC)
  • This (specifically "Policy here should reflect commonsense and actual practice, not the anally-retentive views of a bunch of control freaks") doesn't seem to be at all a civil way to participate in an RfC. --Ohconfucius ¡digame! 08:42, 31 October 2011 (UTC)
  • Michael, can we tone down the rhetoric and depersonalise, please? You've now issued personalised attacks on OC and me. It's unnecessary; let's keep to the issue. Tony (talk) 09:10, 31 October 2011 (UTC)
  • Support following WP:IAR. Its policy that we use some common sense when linking, yes it can be helpful sometimes to use a link twice, and again linking can be overdone. I've wasted quite a bit of time searching backwards through articles looking for the first use of a term.--Salix (talk): 09:41, 31 October 2011 (UTC)
  • Support a more explicit expansion or relaxation of the "rules"/assumptions - per most of the arguments above - even though, as noted, the current guideline is not in fact perhaps as restrictive as some might hope/fear. It currently allows repeat links when the second occurrence is a "long way" - pieces of string, anyone? - from the first, or when one comes in a table or infobox etc. That should almost certainly be extended to cover where a key term turns up, say, both in the lead and the section of the page focused on that issue, even if that is the section directly following (ie not necessarily a "long" way, however defined); also from section to subsequent section, if the term is relevant to both. N-HH talk/edits 13:53, 31 October 2011 (UTC)
    • Problem with that is that many articles may have that appear in most of its sections.∞陣内Jinnai 15:57, 31 October 2011 (UTC)
      • Ah, I'm not saying that it should be compulsory to link a word over and over so that we have one such link in each section, just that it should be allowed, or, at least, not barred. As ever, individual editors can - and should - apply judgment and common sense in each case, eg by looking at how long it has been since a previous link to that term, how relevant/pertinent the term is to the specific section at hand, etc, rather than having restrictive rules about what "can never" or even "must always" be linked. N-HH talk/edits 16:25, 31 October 2011 (UTC)
  • Oppose changing the guideline, which already gives plenty of leeway in saying "There are exceptions to this guideline, including these:..." Dicklyon (talk) 16:58, 31 October 2011 (UTC)
    • The only real leeway states "a long way from the first" which is being interpreted as once per article, and is excluding links after the header - which, as we have seen, is contrary to practice and guidance elsewhere. Obviously the leeway is far from enough. -- cheers, Michael C. Price talk 17:04, 31 October 2011 (UTC)
    • @Dicklyon: I'm not sure the existing wording you cite is very helpful. I submitted an article for GA review, and the reviewer said I had to repeat all the lead links when first found after the lead. I replied, pointing out that WP:LINKS says links should not be repeated unless real far away. The GA reviewer then pointed me to dozens of recent FA articles that repeat the links even when close. This sort of confusion shouldn't be happening. Clearly "repeat links at first occurrence after the lead" is a convention that many top-quality editors are following, so WP:LINKS should at least acknowledge it. --Noleander (talk) 17:23, 31 October 2011 (UTC)
  • I was always under the impression that the repeat links section meant for links within the article body. What I have seen widely practiced is that the lead has a link and so does the first occurence in the body, and then links are not repeated unless they are far away from each other. I very rarely read leads, so it does not make sense to me to have the links only in that section. Karanacs (talk) 19:04, 31 October 2011 (UTC)
Do you support the RfC proposal to enhance WP:LINKS to include a mention of this "widely practiced" convention? --Noleander (talk) 19:19, 31 October 2011 (UTC)
  • Support - Linking once per lead and once per main text (plus eventually one more way down in very long articles) sounds logical and is a very easy to follow guideline - even scripts to check links automatically can use this guideline. However i would not oppose an article with other valid consistant linking methods (see common sense) - except of course for heavy overlinking. GermanJoe (talk) 21:36, 31 October 2011 (UTC)
  • Oppose - multiple links are allowed, there are already five to seven levels of "permission" given
  • # WP:Ignore all rules
  • # This is a guidelineUse common sense in applying it; it will have occasional exceptions."
  • # In general, link only the first occurrence of an item
  • # There are exceptions to this guideline, including these:
  • # Specific examples
    We really don't need to make this guideline longer, no-once objects to sensible linking, and in fact I wonder if we really need the section at all, per WP:COMMONSENSE. Rich Farmbrough, 23:37, 31 October 2011 (UTC).
Sure, the FA linking convention could be construed as consistent with WP:LINKS, as currently written. But only by squeaking past the "avoid repeating links" rule. Look at it this way: WP's top-notch editors at FA have been utilizing particular MOS conventions. Shouldn't the MOS tell novice editors what those FA conventions are? Not as a mandate, but just informational? Shouldn't editors who are nominating articles for GA or FA status be able to go to the MOS and get a clear description of the conventions that are commonly used in FAs? --Noleander (talk) 23:48, 31 October 2011 (UTC)
  • Rich: no-one objects to sensible linking Hah! Reading the last 2-3 years of the talk page should disabuse you of that. And no is suggesting that the guideline should be longer, just replace "once per article" with "one per section". -- cheers, Michael C. Price talk 00:20, 1 November 2011 (UTC)
  • Oppose per Rich. It's allowed at present, and it would not be desirable to turn this into anything stronger, such as a suggestion, recommendation or obligation to link, especially if the word is in the opening of the next section. (But note that I don't unlink such occurrences as a rule, except when there are multiples – 5 or more). --Ohconfucius ¡digame! 01:45, 1 November 2011 (UTC)
    • Perhaps people see this as point-scoring, since the RfC is vague and superseded by the existing text. Michael, one link of the same item in every section is way over the top, even if you might interpret this as "sensible". It would be the ruination of our wikilinking system, and I'm going to stand up for its selective use, which is critical if it's to remain a service to the readers and the project. I have no more time today to engage in this kind of link-in-every-section conversation. Tony (talk) 08:53, 1 November 2011 (UTC)
      • After 2-3 years for some (less than 12 months for me), could we finally get an explanation though of exactly how reducing the number of links available to readers amounts to improving the linking service? That is, unless we assume readers are not capable of exercising their critical functions and making rational choices that are appropriate to their own interests/knowledge levels? I write as someone who is aware of issues around the tyranny of choice and also accepts there is genuine overlinking in a lot of articles. More specifically on point, I personally am opposed to an assumption to link in every section - but equally, this RfC is not redundant, because this guideline does need something more generous, and also clearer, than "it's OK to link again if the second occurrence is a 'long way' off", especially when it comes to lead vs body. N-HH talk/edits 14:38, 1 November 2011 (UTC)
        • N-HH, the usual mantra is the much vaunted "link dilution" - but this concept is invalid, since it is based on commerical click-thru metrics that fail to adjust for the lower likelihood of the text and link near the end of an article being read in the first place (which is correct for calculating click-revenues, but not for assessing readability). I.e. the delinkers here use raw probability instead of the more correct conditional probability. I don't know whether they've ever understand this point, since they never respond to the issue when addressed. -- cheers, Michael C. Price talk 06:57, 2 November 2011 (UTC)
          • As far as my own actions are concerned, a term stands pretty much equal chance of being unlinked whether it is at the beginning of an article or at the end. On the other hand, I am aware there seems to be a greater propensity to link at the beginning of the article, probably due to the repeatlink part of the guideline. --Ohconfucius ¡digame! 07:58, 2 November 2011 (UTC)
            • OC, just as the usual practice, everywhere, is to provide the expanded form of an abbreviation on only the first occurrence in a text, you'd expect the same principle to operate for linking. Michael, you vest me with far greater technical insight that I deserve. I draw on only common sense, not data, for supporting the community standards for selective linking. Tony (talk) 10:22, 2 November 2011 (UTC)
              • That's a convenient memory you have, Tony. No matter, I take it that that's your way of conceding that the oft vaunted data doesn't support your position, despite your many claims in the past that it did. -- cheers, Michael C. Price talk 11:15, 2 November 2011 (UTC)
  • Support Clearly to acknowledge this as an acceptable standard practice and to avoid new editors that read policy too much as a legal framework to take it on themselves to remove such links. Asking for one line or two to be added for this clarity far outweighs the purported issues of CREEP and "it already allows for it". --MASEM (t) 12:40, 2 November 2011 (UTC)
  • Support - I've scrutinized some more FA-quality articles, and the "repeat link at first occurrence after lead" rule is a very common convention. But WP:LINKS, as written now, seems to discourage that: it suggests that links should only be repeated when far apart, yet many FA-quality articles include "near" repetitions, e.g. in the first section following the lead (for one of many FA examples, see Lester Brain). Novice editors, when reading WP:LINKS, should be informed of this convention. I suggest that a bullet be added to WP:REPEATLINKS saying something like "a link may appear at the first occurrence following the lead, even if already linked in the lead" or "when a link is in the lead, the first occurrence following the lead may also be linked". The wording should be crafted to ensure it is merely an option, not a mandate. --Noleander (talk) 15:30, 2 November 2011 (UTC)
  • Support - the current "link only the first occurrence of an item" is not an adequate description of widely accepted good practice, and the MoS should follow usage rather than the other way around. Something a little less prescriptive than "the first occurrence after the lead" might be worthwhile, given the wide support here for a once-per-section (or once-per-large-chunk-of-article) approach. Shimgray | talk | 18:04, 2 November 2011 (UTC)
  • Support. Two arguments in favor of this, from my perspective: (1) If the lead section is intended to be a summary (in some sense) of the entire article, then it can stand alone, independent of the rest of the article. Conversely, the rest of the article might be seen as standing alone, separate from the lead (this is not a strict rule, of course, but is a possible side-effect in reader's minds once they start getting the idea that the lead section merely introduces the topic, and then the rest of the article covers much of the same material in more depth). To the extent that this is true, issues such as linking and footnoting should be treated independently in the lead and the rest of the article. This is, more or less, my view on things. (2) Many readers will not be reading an article from beginning to the end—especially if they are looking for a particular piece of information (which the present article may or may not provide). Such readers might benefit from links further down in the article, even if the same terms are linked higher up. So, yes, the same terms can be linked in the lead and then again later in the article, if editors feel that it is helpful to do so. OTOH, linking "once per section" is almost surely too much; "once per large chunk" (per Shimgray) is better, but, of course, harder to describe in an elegant way... - dcljr (talk) 04:25, 3 November 2011 (UTC)
"If the lead section is intended to be a summary (in some sense) of the entire article, then it can stand alone, independent of the rest of the article. ... the rest of the article might be seen as standing alone".—Put them on separate pages, then; they clearly don't belong on the same page. Tony (talk) 09:14, 3 November 2011 (UTC)
re/ "large chunk", many years ago I picked up "once per screen of text" from somewhere, but while that's a good rule of thumb, it's not one that can be put into a guideline very easily :-) Shimgray | talk | 13:13, 3 November 2011 (UTC)
Tony, please, that just sounds petulant. People do not have to read a whole article if they do not wish or need to. These constant assumptions and even implicit instructions to people as to how they should be reading WP articles, or be able to move around between them via well judged and relevant (as I and most others agree, not indiscriminate) links, are becoming a little much. Readers will often only look at specific sections (many links from other pages link to a section, rather than the whole page); even more commonly, people will often only look at the lead, as a concise - and yes, hopefully standalone - summary of the wider content. With many pages, it's often painful to read much more. And it is, in fact, a long-establised principle that the lead should stand on its own.
Anyway, I think the consensus here is becoming pretty clear on double links in the lead and then a later section, which, as a minimum, can easily be implemented by simply adding "the lead" to the short list at bullet point 2 at wp:repeatlink; there even seems to be some support for more leeway when it comes to repeat links from one section to another, which could be achieved by changing bullet point 1 to say "where the later occurrence is some way from the first, or where it occurs again in a later section in which it has specific relevance or application". N-HH talk/edits 15:09, 3 November 2011 (UTC)
N-HH, you have just called me "petulant". I'm not easily offended, but I urge you not to personalise. My point is that there seemed to be case made that articles should be viewed as a collection of segments. This has no basis in consensus. It would be equally untenable to argue that abbreviations be spelled out at the start of every section. Now, no wording has been proposed, and this RfC is just a collection of rather chaotic views, poorly advertised. No one has given a good reason why an item should be linked again after a short lead, right at the start of the next section: sometimes this will be within five or ten seconds of reading. This is a very poor use of wikilinking. And the current wording allows editorial discretion, anyway. This is indeed an argument for adding bloat. Tony (talk) 13:45, 4 November 2011 (UTC)
No, I said that your specific comment about putting the lead on another page sounded, or came across as, petulant. I'm not being pedantic when I say that not only did I not describe you, personally, as being petulant, but that I was implicitly acknowledging that it probably was not even your intention to sound that way. I just assumed you were not serious about the substantive point, but were just engaging in a bit of hyperbole. Yes, some people read a whole page; others don't. Nor do they have to.
Anyway, I was no more personalising anything than you are when you say people here have not given a "good" reason for any change. Plenty of people have a) supported a change/amendment; and b) have explained clearly why, with perfectly sensible arguments. You disagree, fine, but the actual consensus is quite clear at this RfC (as opposed to the oft-asserted but never-evidenced "consensus" about linking cited by some). As I noted, all it needs is to add one word "lead" to the list at bullet point 2. Surely that will simply add clarity? And it is no more "bloat" than adding one or two links to an article is "linking to the hilt" or a "poor use of wikilinking". N-HH talk/edits 16:54, 4 November 2011 (UTC)
  • Support. I do not believe that we can ever assume that an article is read in its entirety or in a linear fashion. Not when the table of contents allows a reader to easily jump to the section of greatest interest or relevance to what information they seek, and especially when direct section links are not only deployed but becoming even more common as more editors find their specificity the most relevant form of linking. oknazevad (talk) 16:48, 3 November 2011 (UTC)
  • I think that the existing advice is adequate (it is covered by the first bulleted exception), and I oppose any suggestion that the first post-introduction use is the best second link. For example, if the name of a place appears in the lead, is mentioned in passing in the first section after the lead, and is discussed extensively halfway through the article in a section primarily about the place, then the best second link is at the start of that big section, not the first post-lead appearance. But on the general point, more than one link (always in separate sections) is acceptable, exactly like this guideline has always said. WhatamIdoing (talk) 17:32, 5 November 2011 (UTC)
What if a section with a significant use of the term in question is directly after, or comes relatively close to, the lead? Per the strict wording we have now, that would be barred from having a link. In any event, I'm not sure the broad proposal here - or any of the more specific wordings that have been suggested - is mandating that the first post-lead mention has to then be linked as well, especially when it might be just a passing mention; simply that it would be OK to do so (when appropriate in terms of significance, relevance etc), and that a link in the lead does not preclude a link in the main body - whether a long way or not. The point is that the first bullet point doesn't currently deal with the "or not" scenario of course; and hence, per common practice, the guideline does need that additional clarity, surely? N-HH talk/edits 18:50, 5 November 2011 (UTC)
  • Gosh, I wonder if we could rationalise a change in FAC, and instead have 'Featured Sections' on the basis that nobody reads articles in a linear fashion and that every section is therefore independent. ;-) --Ohconfucius ¡digame! 22:50, 8 November 2011 (UTC)
Well exactly: the article, not the section, is the fundamental unit. Sections cannot be the basis for expanding abbreviations and linking unless we change WP so that each section is at a different page. I don't think there's any enthusiasm for that. This discussion is wasting time. Tony (talk) 02:59, 9 November 2011 (UTC)
Indeed, I decide what is and what cannot be, know what is right, and can interpret consensus; as well as drawing absurd conclusions about how what everyone else wants means that every section must be on a different page. More importantly, time is being wasted disagreeing with what a small minority of self-appointed experts - none of whom are presumably actually professionally employed as editors or writers, given the amount of time they spend here - think. Now, sarcasm aside, do we have consensus here to ignore said "experts" and insert one word into the guideline to deal with the lead/repeat issue? N-HH talk/edits 00:06, 12 November 2011 (UTC)
Do it, and hurry before we come to our senses. What ever were we thinking, wasting our time editing here? I could have been wasting my time watching useless TV shows. Dang it. Repent, repent, repent.--JOJ Hutton 00:43, 12 November 2011 (UTC)
  • Oppose the proposed change. The wording we have now is adequate. Generally the lede introduces and links any terms that may be necessary for the understanding of the body of the article. People may indeed read articles in a non-linear fashion, but they will skim the lede first to confirm they have the right article and to get a quick summary, before navigating to some specific section via the contents list. It's wrong to think of the lede as merely a sort of executive summary; it's actually a precursor to the body. That's why we shouldn't normally relink terms from the lede in the body. Colonies Chris (talk) 11:00, 12 November 2011 (UTC)
You might want to open an RfC on changing the opening sentences of wp:lead then ... "The lead serves as an introduction to the article and a summary of its most important aspects. The lead should be able to stand alone as a concise overview". By contrast, I'm not sure where the suggestion that the lead should [only?] link terms necessary to understand the body comes from (however we would define those anyway), in guidelines or common sense and practice. And, while I don't presume to speak for how everyone else might or might not read articles, I can tell you for a fact that I very frequently look at the lead and only the lead; or often only one section, without the lead. We do presumably allow readers to do this; and should, in turn, allow - note not compel - editors to insert a second link subsequent to the lead. N-HH talk/edits 13:09, 12 November 2011 (UTC)
You seem to completely misunderstand or misrepresent everything I wrote. I'm not going to waste my time pointing that out in detail. Try reading my comments again, then retract your own irrelevant response. Colonies Chris (talk) 16:26, 12 November 2011 (UTC)
I completely understood everything you wrote - other than where I clearly signposted one aspect with a question mark - and am not going to waste time pointing out in detail how each part of my response was aimed at each part of what you said (that should be fairly obvious, anyway). Nor, obviously, am I going to retract any part of my comment unless it is precisely explained what I have got "wrong" and why. Whether what you wrote was not what you meant is an issue for you to address, not me. N-HH talk/edits 16:54, 12 November 2011 (UTC)
  • Oppose. I think excessive and reflex use of linking is getting out of hand. We should resist the idea that if a capability exists, it's to be used at every opportunity. I am reminded of the reflex use of centred headings, in amateurish wordprocessing. Novices learn how to do it, and then seem to consider the new trick an automatic source of merit and kudos. Never mind readability. Linking is like that. It is too often a substitute for thoughtful exposition in the present text. I oppose any shift from the status quo that would encourage this excess. I strongly favour a clean text that stands up on its own. One well-placed link is better than three duplicates that are thrown in "just in case". As the discussion above shows, there is already enough flexibility to cover any special need that might arise. Leave it alone. NoeticaTea? 12:58, 12 November 2011 (UTC) Note: I did not come to this RFC to have my opinion misread and misinterpreted, or to endure the abuse of those who take a view different from my own. I have spoken of "novices" here; so have others. I have not been disrespectful to any participant here, and I do not expect to be treated with disrespect myself. This is my last involvement in the discussion until I see retraction and apologies. The RFC is tainted by such abuse; accordingly no consensus for change can be properly established. NoeticaTea? 14:16, 12 November 2011 (UTC)
Have you read the above discussion and taken on board what the issues are? The status quo - in terms of practice and FAC - is precisely to encourage a second link that puts a link to a relevant term in the body as well as the lead. The point is to make this guideline reflect that. The point is very much not to change the status quo, to promote the insertion of hundreds of links or even necessarily to compel the introduction of a second link in the body. It is to be explicit in saying that the option exists, and that editors can use, you know, their judgment in how and when to exercise it. Also, this patronising tone - "novices", "reflex", "[supposed] kudos", "amateurish" etc - comes across as a bid to set some up as "experts" and others as ignorant about how to link properly. Can we see your certificate of expertness? N-HH talk/edits 13:09, 12 November 2011 (UTC)
@Noetica - The fact is that most FA-quality articles do repeat the link at the first occurrence after the lead (including today's article Tumbler Ridge, British Columbia). The point of this RfC is to help novice editors reconcile what they see in FA articles (repeating links) with what this MOS says (don't repeat). Right now, novices see two contradictory practices. Don't you think we should clarify that apparent contradiction (without mandating one way or another) for editors? --Noleander (talk) 13:38, 12 November 2011 (UTC)
I think I have seen enough. At the opening, there was some berating of "anally-retentive views of a bunch of control freaks"[1]. Then we have "small minority of self-appointed experts"[2]. Now, we have what appears to be some persistent riposting[3]. So much for the alleged intolerance (or whatever the exact word used was) of the "small minority of self-appointed experts [sic]". Notwithstanding arguments about non-linearity, NHH and the other supporters failed to demonstrate how subsequent links are deterred by the current guideline that they need to make such explicit, but have also abjectly failed to indicate exactly how such 'permission' they seek will not translate into systematic linking of the first occurrence of a 'contextually relevant' word in each section. --Ohconfucius ¡digame! 13:40, 12 November 2011 (UTC)
As Noleander says, this is a no-brainer. We add one word to the guideline, and we get clarity and consistency. We have clear consensus for that, even if one or two editors - who are indeed "self-appointed experts"; if you don't want to be referred to as such, don't act that way by arguing other people don't know what they are doing or haven't "proved" something or other - disagree. That is your right. However, there is, contrary to some suggestions, no "right" or "wrong" way to link, or to advise on linking, which those experts can divine for the rest of us and impose on the community. That is what I have kicked against ever since I came across the broader linking issue. In reality, all we have at the end of the day is the individual preference and guesswork of thousands of readers and editors, expressed and mediated through practice and consensus. Btw I'm not sure why I or anyone else needs to show either of the two things you mention, even if it were possible to do so. So our supposed abject failure is neither here nor there. N-HH talk/edits 13:55, 12 November 2011 (UTC)
If you don't want to be referred to be described as an "anally-retentive bunch of control freaks" (sorry a "small minority of self-appointed experts") then stop behaving like one – Very droll indeed. "We add one word to the guideline, and we get clarity and consistency." Pray, what is that one word? Oh, and by "consistency", presumably you are talking about repeat linking the 'relevant' word(s) in each and every section? --Ohconfucius ¡digame! 14:34, 12 November 2011 (UTC)


Sorry, there's no consensus for any specific wording. I made that point some time ago. All we have is a set of mostly loose comments, only to be expected when an RfC is framed so loosely. Tony (talk) 14:53, 12 November 2011 (UTC)

I don't think anyone has said that there is, yet. The RfC question is as follows - "Should WP:LINKS be modified (probably in the WP:REPEATLINK section) to acknowledge this as a permissible (but not necessarily recommended) practice?" So, no it does not ask for approval for a specific wording, but the RfC proposer did respond to that point, also some time ago - "I deliberately left that out of the RfC description, because a particular wording might narrow the discussion down too far ... that kind of detail can wait until if/when this RfC consensus turns out to be 'Support'". That seems reasonable enough to me, and we now clearly have consensus - if not unanimity - in favour of that general proposition. Hence, we can now move to working out what wording satisfies that consensus; and if there's a simple tweak to the guidelines that uncontroversially implements the consensus, we don't even need to waste more time wrangling over the precise wording. As for my primary suggestion, it was - "simply adding 'the lead' to the short list at bullet point 2 at wp:repeatlink". So, apologies, it was two words, including the definite article; and "consistency", btw, refers to what would then be the consistency between editor and FAC practice, and the guideline here. Anyway, that's certainly one option, and seems to be the simplest and quickest. I also detect the rumblings of a bid to discredit this RfC from those who disagree with its conclusions (see also Noetica's post above). I hope that's just me being paranoid; I can't see the benefit to anyone in going down that road. N-HH talk/edits 16:02, 12 November 2011 (UTC)
Sadly, N-HH, I don't think you're being paranoid. 2 delinkers have indicated their intention of ignoring the consensus that has emerged. The consensus is clearly to specifically permit duplicate linking immediately following the lead. As some others have pointed out the only question is how much to relax the guidelines in general. My own proposal would be to allow linking in sections, regardless of linking elsewhere, simply because readers can be navigated directly to sections from elsewhere. -- cheers, Michael C. Price talk 16:37, 12 November 2011 (UTC)
"Duplicate linking immediately following the lead" is very different from "once per section" which you suggested was the proposed outcome in response to my comment above. I suspect that the RFC would have had a better chance of achieving clarity if it had said change to wording to "links may be repeated one per section" or "links may be duplicated immediately following the lead". I also think that neither is what is wanted. You may have better luck digging through history to see if there was a version that said something about links not being repeated "close together" - that may stand a good chance of getting more uniform support at least for the idea, if not for an explicit change. Rich Farmbrough, 01:08, 13 November 2011 (UTC).
Yes, Rich, I know the two suggestions are different, which is why I said "my own proposal". As for clarity, the RFC clearly indicates linking immediately after the lead is permissable and that the MOS guide needs amending - if the results were not clear then we would not have the declared intention of two delinkers to ignore the RFC consensus. As for the history, there was a stable version that said that links should not be repeated within a section; stable, that is, until the present delinking clique started their crusade, declared their views consensus and ignored all other opinions. And engaged in tactics such as archiving active talk threads. -- cheers, Michael C. Price talk 08:58, 13 November 2011 (UTC)
The current guidelines allow editors to judge if a duplicate is useful to the reader. No one here has explicitly agreed to a situation where an item is linked in a one- or two-sentence lead (most WP are stubs or not much bigger), then straight away linked again, within five seconds of reading. With an infobox link (explicitly permitted), that makes three links at the top of an article. This RfC is a mess, I'm afraid, and provides no endorsement of one person's wording, invented after the fact. Tony (talk) 12:45, 13 November 2011 (UTC)
And not to mention that it got very personal just as the ball on this started rolling. --Ohconfucius ¡digame! 13:04, 13 November 2011 (UTC)
  • Question to the opposers: I think the status quo situation, in which the guideline and the actual practice in a s***load of articles (including many featured ones) disagree with each other, is dysfunctional; hence, I hope that the people who support the current text of the guideline would be willing to roll up their sleeves and fix all those articles, removing links to articles already linked to in the lead unless there's a good reason to keep a particular duplicate link, rather than hoping that a guideline that hasn't been followed thus far will somehow magically be followed in the foreseeable future. Amirite? ― A. di M.​  18:57, 13 November 2011 (UTC)
Oh god, don't go putting ideas like that into people's heads; I'm sure there are several that would leap at the chance to do exactly that. Although, were they to, they would of course be in clear breach of the consensus established here, as well as the consensus already set by common practice. There's quite enough mass delinking going on currently (and the fact that there's no consensus in favour of much of it - despite assertions, never evidenced, that there is - or that it takes a lot of time to put into effect has been no bar to that). N-HH talk/edits 20:00, 13 November 2011 (UTC)
A di M, what would be FAR more logical would be to remove the link in the lead and put it (or retain it) in the opening section. Twice is silly; and the lead is the very last place a reader would want to divert to another article, isn't it? We're not here to cater for this ten-second browser kind of notional (I'd say fictional) reader, who wants to skip from one article to the next after reading the opening sentence. The lead looks down into its article. Tony (talk) 02:28, 14 November 2011 (UTC)
Ah, this “divert to” thing again, as if tabbed browsing hadn't been supported by pretty much all major browsers for several years now. Plus, it's nearly obvious that in most cases more people read the lead than a section below it. (A principle I once read somewhere, though I can't find it right now, is that an encyclopaedia made up by the leads of Wikipedia articles should make sense; and the point of the ToC is that someone can read the lead and Section 2 without reading Section 1, so if a term is repeated in all three, linking it only in the lead caters such readers, whereas linking it only in Section 1 doesn't.) And isn't the fact that you can use articles as stepping stones the very reason we don't like ‘chained’ links in the same sentence? If the top screenful of Boston doesn't link to Massachusetts, saying “There's no point in linking both Boston and Massachusetts in the first sentence because blah blah blah” isn't a valid argument.
That said, there are some cases where the best place to link to an article is not the first mention, but they are a minority. ― A. di M.​  11:45, 14 November 2011 (UTC)
I already do that when I come across instances of links repeated several times in the lede – they are surprisingly common. It doesn't surprise me to find the same word(s) linked many more times in the body. --Ohconfucius ¡digame! 02:39, 14 November 2011 (UTC)
  • Support. I like blue links. (Blue is my favorite Wikipedia color.) They should at minimum be repeated when they are far enough apart that the repeated link doesn't appear on the screen at the same time as the previous one. (Of course that varies with the individual's screen, so editors could reasonably differ on this.) Red links should only appear once in an article, and should not be repeated. (I definitely do not like red links! Somebody should either write an article to make them a beautiful blue, or if it's not notable, delete the red link!) --Robert.Allen (talk) 08:51, 14 November 2011 (UTC)
  • Admittedly, blue is a much less traumatic colour than red: Blue is the colour of cool, calming water, red is the colour of molten engulfing lava. I also like seeing blue on the screen, but only small blotches of it suffice to give that calming effect. The reality is, red links are more useful than blue links, and although people seem to have an aversion to them to the extent of liberally removing them on sight, they remind us of the job at hand, which is really to "Build the Web". Red is what motivates me here on WP. Ironically, the parts already built don't need so much attention drawn to them. They need only to be linked once. --Ohconfucius ¡digame! 11:07, 14 November 2011 (UTC)
  • Comment - I posted a request at WP:AN asking for an admin to read this RfC, see if there was consensus, and close it. Over a week has gone by with no new activity. However, the full 30 day period recommended for RfCs has not elapsed, so if anyone wants to let it run another few days, that is fine also. --Noleander (talk) 15:40, 25 November 2011 (UTC)
I remind editors that this RfC was poorly conceived if the intention was to change the wording of the guideline. It has proposed no new wording, and the wording itself needs to be the subject of consensus if it's to be changed. The responses of editors here, instead, cover a range of opinions. I suggest another RfC proposing actual wording be held if anyone is thinking of changing the guideline. Tony (talk) 15:45, 25 November 2011 (UTC)
Sure, the change could proceed in two steps. The thrust of this RfC was "Should the guideline be improved to acknowledge the practices often used in FA articles?". If the closing admin finds that the above discussion resulted in the answer "Yes", we can then start a discussion of the precise wording, either with an RfC or just an ordinary discussion on this Talk page. --Noleander (talk) 15:49, 25 November 2011 (UTC)
  • Support per Michael C. Price. Long overdue, and actually not going far enough. As usual, when we got rid of the bizarre overlinking of dates etc. the pendulum swung too far in the opposite direction. Hans Adler 16:02, 25 November 2011 (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.

Table of non-obscure units

Latest comment: 13 years ago11 comments5 people in discussion

I don't feel that the exact choice of which units to list has been discussed enough for such a table to be shown. For now, I'd just remove the table. (Also, what happened to the point about conversions? The degree Fahrenheit is very obscure outside the US but not obscure at all inside it, and vice versa for the degree Celsius, so you want to link degrees Celsius if used alone, and degrees Fahrenheit if used alone, but you don't want to link them if used together for the same temperature, as in 20 °C (68 °F) or 68 °F (20 °C).) ― A. di M.​  21:20, 3 November 2011 (UTC)

Agree with removal. The whole idea that some units are well-known universally ignores the multiculturalism of the Anglophone world. -- cheers, Michael C. Price talk 05:16, 4 November 2011 (UTC)
Well, a few units are known pretty much universally: year is one of the most commons nouns in English. :-) ― A. di M.​  16:53, 4 November 2011 (UTC)
True, perhaps the table does have too much information (particularly if not discussed in this Talk page before it was created). Rather than deleting the whole thing, perhaps just replace it with some prose that gives a few examples. That section already has the sentence; "... the troy ounce or bushel, the candela or mho might be considered obscure." Perhaps add a second sentence like "units that are relatively common and generally don't need to be linked include watt, volt, hertz, bit, byte, hour, minute, and second." And maybe a third sentence: "Other units may be obscure in some countries, but well-known in others (such as metric system units, which are not well known in the US) and so linking them may be useful." Or something like that. --Noleander (talk) 13:25, 4 November 2011 (UTC)
Wouldn't it be better to discuss each item, point by point, here, before throwing it out? I'm keen that exceptions to general rules be listed in a table or appendix at MOSNUM: there is actually chaos out there WRT some units, and a few edit-wars and fights about them, I believe. This is what our style guides are meant to minimise. A di M, the anglophone world is about as multicultural as plasma. English seeps into everything like a kind of mould, and as you know, has come to dominate the world. Tony (talk) 13:36, 4 November 2011 (UTC)
I think it makes more sense to discuss each item, point by point, before adding it. I don't perfectly remember the details of the story, but the whole table was introduced after a hurried discussion at WT:MOSNUM which AFAICT wasn't even advertised here. (And I like Noleander's idea of a few examples on which everyone would agree, but adding “unless a conversion is present” at the end.) ― A. di M.​  16:43, 4 November 2011 (UTC)
As for “the anglophone world is about as multicultural as plasma. English seeps into everything like a kind of mould, and as you know, has come to dominate the world”, I completely agree, but what the hell does that have to do with this, exactly? ― A. di M.​  16:46, 4 November 2011 (UTC)
It means that we shouldn't make assumptions about the background of our readers - so leave the links in. -- cheers, Michael C. Price talk 16:48, 4 November 2011 (UTC)
Knowing Tony, that's likely not what he meant. ― A. di M.​  16:53, 4 November 2011 (UTC)
Yes, no doubt he has a rationalisation about why anglophone multiculturalism means we should remove links. -- cheers, Michael C. Price talk 16:57, 4 November 2011 (UTC)
I thought this table was a bit ropey, that's why I amended it a little, and added some text. Maybe it could be spun off into a sub-page, to avoid cluttering, and extended appropriately. Rich Farmbrough, 01:14, 13 November 2011 (UTC).
Latest comment: 12 years ago12 comments10 people in discussion

The MOS says the following:

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.

It's not immediately obvious why this would be deemed good practice, rather than bad practice. Furthermore, following this rubric - especially the end of it - isn't helpful to the reader. I'm fairly sure the question "why?" is an FAQ here (though I can't see a list of FAQs) - but why do we have this instruction?

Assuming there is a good reason, perhaps it could be included in brief in the actual MOS text. --Dweller (talk) 16:09, 10 November 2011 (UTC)

Perhaps the reason for that guidance is that a link within a quote give an erroneous impression to the reader, such as (1) the blue color may imply some emphasis (by the original speaker) on that particular word/phrase; (2) there could be OR issues if the editor chooses a target article that is not 100% identical to the quoted word (e.g if the quoted word is "foo" but the target article is foobar); (3) if there are a few articles with similar names (disambig) there is some OR if the editor picks one of several for the target. But that is all just speculation on my part. Personally, I see no big problem with including links within quotes. --Noleander (talk) 17:17, 10 November 2011 (UTC)
I see no reason why links can't be included within quotes either. -- cheers, Michael C. Price talk 07:16, 12 November 2011 (UTC)
Yes this is slightly dodgy. Advice to take special care linking items in quotes for the type of reason Noleander outlines would make more sense. Rich Farmbrough, 01:16, 13 November 2011 (UTC).
The reasons for not linking within quotations are well-founded, and should be obvious — but a few additional words could be added to the first part of the sentence to firmly explain why. We editors are already burdened with taking special care to not misquote, or quote out of context, and we don't need to be given additional special license to perform mindreading on the person we quote, as well. A quoted persons words are immutable, whereas the Wikipedia articles to which we link are not. Even when an editor links to an 'identical' quoted word, there is no guarantee that our Wikipedia article conveys the same meaning as that intended by the quoted speaker. There is further no guarantee that the meaning conveyed by our article won't change two days from now. When we quote, we are to convey what the source says, not what we personally think the source means; we have crossed that line when we start Wiki-linking the words and phrases within a quote. Xenophrenic (talk) 17:55, 13 November 2011 (UTC)
Why is it a sin to link "a far-away country" in Chamberlain's speech? It provides context. Of course judgement should be used, and care taken, but an outright ban is silly. -- cheers, Michael C. Price talk 18:06, 13 November 2011 (UTC)
WP:EGG. Much better is “a far-away country [Czechslovakia]”. ― A. di M.​  18:47, 13 November 2011 (UTC)
I agree with the guideline, but I also agree that it should state its rationale. ― A. di M.​  18:51, 13 November 2011 (UTC)
I strongly support this sentence. The only problem I see with it is that it is not strong enough. Some editors consider following it completely optional, so they are free to insert original research and POV statements into quotations. Any literal quotation that comes with a link that is not due to the original author is a misrepresentation of what the author actually said. By doing so we always put the words into a new context which is not the original. Sometimes it comes close enough to the context intended by the original author, but much more often it does not. Editors with poor judgement or a poor sense for linguistic nuances have trouble understanding this and then insist they can simply ignore this sentence. This problem needs fixing. Hans Adler 16:07, 25 November 2011 (UTC)
Yes, context can always be stated in other ways, as suggested. Having been a former long-time editor fo Falun Gong articles, support the current wording. I am well aware of how this device has been used in situations to give an emphasis where it was certainly, in my opinion, undue. Of course, the risk of introduction of original research is ever present if this practice is allowed. --Ohconfucius ¡digame! 00:43, 26 November 2011 (UTC)
I agree with Rich Farmborough. We went from a complete ban on linking within quotations to a slightly looser guideline several years ago, which seemed fair enough. But don't you all think it's wise to flag to editors that a little more caution than usual is required in the judgement of whether to link an item in someone else's text, as opposed to WP's narrative? Among the issues are that the "owner" of the text didn't link to a particular article in their text, that it introduces another variable given that a quotation-linked item can change over time to represent an angle or view incompatible with the original writer's/speaker's intended meaning. I'm not for a total ban, but editors should be reminded of these matters, I think. Tony (talk) 08:37, 26 November 2011 (UTC)
I can see that, but would still like more clarity and more leniency. Yes, linking can imply things that the quoted party would not have implied (this is a WP:NOR problem, already covered by a policy), and can be manipulated to seem like improper emphasis (already covered by WP:NPOV). The main problem with not linking in quotations when linking would be helpful is that the passage is either not as encyclopedically useful as it should be (maybe even confusing), or it must be followed with explanatory verbiage which is redundant and tedious, and worse yet is often even more prone to WP:NOR and WP:NPOV problems than the simple links would have been. — SMcCandlish Talk⇒ ʕ(Õلō)ˀ Contribs. 19:45, 7 January 2012 (UTC)

Arthur's edit

Latest comment: 12 years ago2 comments2 people in discussion

Arthur and I have discussed his recent edit concerning timeline articles. I've reverted his change to the guideline pending discussion here of just what was decided at the big RfC on dates. At the very least, a link to the RfC is required; and I distinctly remember that Arthur conceded to Colonies Chris early this year that timeline articles are not part of the "exemption" granted to articles on years, months, days, etc. Input would be appreciated. Tony (talk) 08:27, 26 November 2011 (UTC)

The quote you found referred to such articles as November 2008 in sports; I withdrew the complaint after looking closely at the article and found no standard of whether dates should be linked within the article. That does not apply to articles written to a standard, such as 2011 in the United States. — Arthur Rubin (talk) 09:21, 26 November 2011 (UTC)

Timeline

Latest comment: 12 years ago27 comments7 people in discussion

There is clear consensus that the chronological linking guidelines do not apply to day-of-year, year, decade, century, millennium, and some other articles. I thought it was all timeline articles, but I may have been wrong. However, some indication of that consensus should be here, regardless. Specifically, 2011 in the United States has been linked until recently, so there needs to be some discussion before it is unlinked. I believe 2000s in sports has also been linked, and not just to 2000 in baseball, etc. — Arthur Rubin (talk) 08:37, 26 November 2011 (UTC)

Actually, there is no consensus at all. The community spoke loudly and clearly at that turning-point RfC in 2009. I see no reason its voice should be nullified right now. Tony (talk) 09:44, 26 November 2011 (UTC)
Actually, you're quite wrong. The community spoke loudly, and clearly indicated that there were exceptions for types of articles, and did not identify the exceptions. — Arthur Rubin (talk) 09:56, 26 November 2011 (UTC)
According to my recollection, "and some other articles" was deliberately vague because there was never any agreement how to properly define it. If Arthur thinks that timeline articles was to be included, he should demonstrate that the consensus existed for it then, or that it exists now, before he goes about changing the guideline. Thanks. --Ohconfucius ¡digame! 12:41, 26 November 2011 (UTC)
The guideline already says, “Intrinsically chronological articles (1789, January, and 1940s) may themselves contain linked chronological items.” BTW, is there any reason for this section and the one before to be separate? ― A. di M.​  11:19, 26 November 2011 (UTC)
Arthur, as Colonies Chris said to you just before you bullied him with threats of blocking: "Please explain to me why you think a reader who's interested to find out about events expected to happen in 2015 would benefit from a link to the events of, say, March 20, in some random set of completely different years. The only date link that's even arguably relevant, albeit remotely, is '21st century': the month-day links are of no value whatever." The same might be said of linking 3 March in a timeline article. I suppose this is why the community was so dismissive of the idea. Tony (talk) 13:17, 26 November 2011 (UTC)
That is not the consensus. "Intrinsically chronological articles" are exempt, and there has never been any consensus that 2011 in the United States is not "intrinsically chronological". I can't agree with the anon that there should be articles January 1 in the United States/1 January in the United Kingdom, but, if there were, keeping the date links in place would be the easiest way to adjust them later. As it stands, I do not have a bot that links only dates in line headers, and I doubt that should a bot could easily be written; yet another reason for retaining the date links if there is any doubt.
Finally, if you are going to accuse me of bullying, it would be appropriate to note that this entire episode is a good example of the big lie; certain editors asserted, without evidence, that consensus was against date linking. Sometime later, after a number of RfCs which could not possibly have produced a result which would not be interpreted as consensus against date linking, one was finally produced which could not possibly have a produced a result (other than no consensus) other than being against date linking. I agree it did result in a consensus against date linking, but, if you want to revisit the exceptions, I would propose an RfC which could produce a different result. — Arthur Rubin (talk) 16:36, 26 November 2011 (UTC)

Arthur, the principle is clear: Month-and-day articles (February 24 and 10 July) should not be linked unless their content is germane (relevant and appropriate) to the subject.

Please read that above principle again; three times if necessary to fully appreciate what it is driving at.

Timelines and anything else you want to write about are no exception to this principle. If a reader is in our Timeline of antibiotics article, which begins:

• 1910 - Arsphenamine
• 1912 - Neosalvarsan
• 1935 - Prontosil

…all we do is confuse readers, waste their time following extraneous links, and make them fear even clicking on links when we leave time-wasting blue land mines that are totally irrelevant. If the article is directed to a general-interest readership that came to read up on antibiotics, precious few readers will be impressed when they get to 1910 where they will find mind-boggling jewels like “July 4 – African-American boxer Jack Johnson defeats American boxer James J. Jeffries in a heavyweight boxing match, sparking race riots across the United States.” In fact, the 1910 article doesn’t even contain the word antibiotic.

I remember the first time I clicked one of these trivia links. I was interested in scientific fiascos and our article on it had something like The 1967 announcement of polywater. I thought, “Wow, I can go to a detailed account of that announcement!” All I learned was “Don’t click those fucking links.” It was clear that Wikipedia had fallen victim to a legacy of its early days when there were far fewer articles and wikipedians hyperlinked any and everything. Now…

We’ve been through all this this before. Once readers are reading one of the intrinsically chronological articles (which are lists of unrelated trivia) such as 1910, then it is OK for those articles to themselves contain still more chronological links such as July 4.

There is one notable exception to this “chronological trivia can lead to more chronological trivia” rule. I am pleased to announce that our Trivia article is, IMHO, completely compliant with our principles on linking. It has this:

On September 13, 1965, four Columbia students appeared on the TV quiz show I've Got a Secret and competed in a trivia contest with the show's regular panelists.

Boy, oh boy, are our visiting readers who are interested in trivia in for a treat. Greg L (talk) 19:48, 26 November 2011 (UTC)

Wrong, as usual. As even you must be aware, "inherently intrinsically chronological articles" are exempt from unlinking. The only question is whether 2011 in the United States is "inherently chronological" (arguable, but there is not a clear consensus, except by the regular editors of the article), and whether the regular editors want the links there (not arguable; the answer is yes). — Arthur Rubin (talk) 00:40, 27 November 2011 (UTC)
Awe, you are so charming when you’re at your best. Juvenile responses like that do not help your case. If your position is now clear, that’s fine, but your addition (…these guidelines do not apply to timeline articles…) (∆ edit here) was incorrect policy if interpreted as it was written and was incorrect if you meant something other than what you wrote. Our Timeline of antibiotics article is an example of a timeline article. Please exercise more care when unilaterally changing Wikipedia’s guidelines; writing one thing and meaning another happens to us all, but getting all pissy and attacking others when they revert your mess is another. Is this response something even you can understand? ;-) Greg L (talk) 00:55, 27 November 2011 (UTC)
That was partially my mistake. I didn't (and apparently, neither did the unlinkers) see the note that the guideline does not apply to "intrinsically chronological articles". I don't think that applies to timeline of antibiotics, but it does seem to apply to 2011 in the United States or 20th century in antibiotics (if such an article existed). — Arthur Rubin (talk) 01:13, 27 November 2011 (UTC)
I consider myself an “agnostic” with regard to those date links in 2011 in the United States. If it was 2011 in baseball, its very targeted subject matter (baseball) makes it inappropriate, IMO, to have links to trivia that is nearly 100% unrelated to baseball. And, after writing that baseball-related article title in brackets and seeing it is a real article—and reading it—I see that the dates in it are not linked. That’s all good and well and proper, IMO.

But I can also see why many would argue that 2011 in the United States is so sufficiently broad (the U.S.), that readers might be interested in following links to see what else was happening in the world at that time. I would have to see persuasive, rational, thoughtful arguments both ways to come down one way or another for articles like 2011 in the United States. All those links there now don’t exactly shock my wiki-conscience.

But having linked dates in, say, 20th century in antibiotics would be inappropriate, IMO, because of its targeted focus on a very specific (narrow) subject (antibiotics).

Having written all that, I would propose something like this for debate: For “[Time period] in [subject]” articles such as 2011 in the United States, where the subject matter (United States) is very broad and and is not highly specific (such as Timeline of antibiotics or 2011 in baseball), chronological items (dates, months, and years) may be linked. The underlying principle behind this guideline is that many readers of articles covering very broad subject matter would be interested in following links to see what else was happening in the world at that time. I could support this, which has the virtue of reflecting the current reality. We could benefit from having the current reality memorialized in our guidelines so we don’t have further editwarring over it.

This proposed guideline would also cement an important principle on the flip side of this issue; an article like 20th century in antibiotics would not properly have linked dates, and the reason has nothing to do with the fact that it has “20th century” in it, but is because readers of such highly focused subject matter would typically have no interest in going to trivia articles totally unrelated to the subject at hand. Greg L (talk) 02:20, 27 November 2011 (UTC)

Actually, that seems reasonable, with the additional notes that, for example, 2011 in baseball could rationally link to [[1932 in baseball|1932]] or, rarely, [[2011 in football|football]], when referring to an athlete known for more than one sport; and some discussion related to where the breakdown is between 2011 in the United States and 2011 in Tuvalu (where we might agree that the one potential entry shouldn't have links). It may need work, and possibly another <censored> RfC. — Arthur Rubin (talk) 02:48, 27 November 2011 (UTC)
I’m pleased that you see that as a potentially acceptable proposal to work on. There is no point letting details get in the way of things, but I’ve long thought that aliasing a link to the point that it is at great risk of appearing to be something it is not or which makes it ambiguous, such as this code:

Willie Mays won the [[Negro League World Series]] in [[1948 in baseball|1948]]

to produce this:

Willie Mays won the Negro League World Series in 1948…

…does our readers—and even our wikipedians—a disservice. Our readers are accustomed to something that reads “1948” as taking them to the trivia article with that exact title. Why would any wikipedian want to write informative, useful text that masquerades as something that isn’t nearly as useful? Were it me, I’d write it this way:

In 1948, Willie Mays won the Negro League World Series (see 1948 in baseball ).

Greg L (talk) 03:39, 27 November 2011 (UTC)

I see 2011 in baseball does not have any year links of any sort, but, for instance, in the deaths section, it would not be unreasonable to have, in deaths:
  • February 30 – Famous Player, had league doubles record in [[1960 in baseball|1960]] for the (insert team name here) (b. [[1939 in baseball|1939]])
I can't say I'd recommend it, but it shouldn't be banned. That's not related to this guideline, so let's let it go. We still need some guidelines for "sufficiently broad". (And I still think an RfC might be needed to avoid trouble; if we can agree on the wording, there should be few who would object that the wording is unfair.) — Arthur Rubin (talk) 03:55, 27 November 2011 (UTC)
I can’t agree with you, Arthur, about linking “February 30" in such an article. Your example squarely flouts the basic principle of Month-and-day articles (February 24 and 10 July) should not be linked unless their content is germane (relevant and appropriate) to the subject. Too few readers interested in baseball are going to find value in being taken to articles that have nothing whatsoever to do with baseball. Your example link (linking “February 30”) is a treatise with such information as how the Soviet calendar treats “Feb. 30” differently than the Swedish calendar. That sort of tangent is not *germane* to baseball and excessively stretches the Chain Of Discovery®™© (Baseball → Famous Player → Player died → Date of death is mentioned [apply philosophy of “hyperlink early and often” here] → Soviet Calendar treats “Feb 30” differently than Swedish calendar). Greg L (talk) 23:25, 27 November 2011 (UTC)
On the other hand, lots of readers on seeing “February 30” will assume it's a typo, and on following the link they will realize that maybe it is not. ― A. di M.​  23:45, 27 November 2011 (UTC)
Wow… Sad, but true. It could be one of those Mystery Easter Eggs where something useful masquerads as something to avoid like the plague (is it [[1948 in baseball|1948]] or is it [[1948]]??). It elicits a reaction of “Did wikipedians accidentally screw up or accidentally-on-purpose screw up?” Greg L (talk) 00:27, 28 November 2011 (UTC)
  • What, pray tell, are the navboxes in these articles for? Washing blue over everything in each entry weakens the wikilinking system. I still don't understand why May 3 is relevant to a timeline of any sort. Tony (talk) 09:27, 27 November 2011 (UTC)
  • Policy idea - Date linking is usually acceptable for articles that have a date/day/year in their title. It is only acceptable in other articles if the destination page is immediately relevant to the topic, but in most cases, this should be avoided. Is this where the discussion is going? D O N D E groovily Talk to me 03:12, 1 December 2011 (UTC)
    I'd be OK with that, but I don't think the unlinking WP:CABAL would go along. I also don't think it was exactly what I was proposing. — Arthur Rubin (talk) 09:19, 1 December 2011 (UTC)
    The ArbCom clerk in charge of that enormous RfC back in 2009 inserted these clauses into MOSNUM as a direct result of the community's very significant consensus. It was a direct reflection of consensus, and I see no reason to detract from that. This is why I repeated the question that was asked at that RfC, thus far unanswered at this page. Let's not waste time on this. BTW, in case you think I have some inherent bias against year, century, date articles (i.e., the intrinsic ones), note that I've added my clear support in a number of venues to the continued presence of On this day at the main page. OTD contains I don't know how many year and day-and-month links, every day, exposed to millions of views. One should not tempt providence: OTD scored rather poor support at the RfC a few months ago on what to keep on the main page and what to dump. Tony (talk) 11:54, 1 December 2011 (UTC)
    @Dondegroovily: That's way too broad. I can see no good reason why a date should be more likely to be linked in (say) 2009 L'Aquila earthquake than in (say) Ultra-high-energy cosmic ray just because the former contains 2009 in its title. ― A. di M.​  18:44, 1 December 2011 (UTC)
    Good point. We would need to exclude articles that use dates only as disambiguation (like disaster and elections articles). That makes things more complicated. D O N D E groovily Talk to me 01:55, 2 December 2011 (UTC)
    And there's no in-principle difference between "History of X" and "Timeline of X". We don't allow year or day-and-month linking in the first without very good reason; nor should we in the second. Tony (talk) 14:23, 2 December 2011 (UTC)
Knowing that we shouldn't be linking years without good reason, it would seem to me that if you have a History of X or Timeline of X article, presented in a non-prose format (table, list, whatever), and there exists a series of "YYYY in Z" (doesn't necessarily have to be years, could be decades, whatever, if it is a longer-term subject) where Z is the immediately obvious parent topic of X, then having links from the years give to the specific "YYYY in Z" article makes sense for tracking the influence of topic X on and by the larger field Z. But this has to be an obvious, intuitive relationship. In no case would I link to a bare year article (barring unless we're talking a pure calender-based discussion), and, for example, if we're talking a development in medicine that just happened to occur on US soil, there would be no point in linking that to the year article in US history - neither are non-obvious connections. But linking year dates from within History of antibotics to the appropriate YYYY in Medicine is exactly the type of application where these links make the most sense. --MASEM (t) 14:47, 2 December 2011 (UTC)
  • I agree with A. di M.’s 18:44, 1 December 2011 post.​ Just because a Supernova 1987A contains a date doesn’t take the pure trivia found in our 1987 article and magically make it “germane” to “supernovas”. There is too much wikilawyering going on here by a few editors that seems to amount to nothing more than efforts to probe for logical wiki‑loopholes. The most modest application of WP:COMMONSENSE is all we need here; in no dictionary will one find “profoundly irrelevant” as a definition of “germane”. The colossal effort put into the 2009 RfC and its abundantly clear consensus is in no need of being turned on its head. Greg L (talk) 02:09, 4 December 2011 (UTC)

Month-and-day

Latest comment: 12 years ago3 comments3 people in discussion

A. di M.: Why didn't I think of that as an alternative to month–day? Tony (talk) 13:20, 26 November 2011 (UTC)

(I had considered day-of-the-year too, but that sounded a bit too clunky.) ― A. di M.​  16:12, 26 November 2011 (UTC)
Thanks. I wish I had thought of that. Now, on to WP:DOY.... — Arthur Rubin (talk) 00:41, 27 November 2011 (UTC)
Latest comment: 12 years ago2 comments1 person in discussion

SSilver: you just changed WP:REPEATLINK to read: "the same word or phrase should be linked only once in the lead and once in the body ...". I believe that is not consistent with the RfC above in this Talk page, which concluded that such double-linking was optional, and that this guideline should not recommend (nor discourage) such double linking. I think the wording that was there before ("... it may be useful ...") shows the optional nature of the double links. There are probably lots of ways to word it, but "should be linked .." is not consistent the the RfC outcome. --Noleander (talk) 22:40, 29 November 2011 (UTC)

I've changed the wording to "... but if helpful for readers, links may be repeated...", which I believe is more consistent with the RfC consensus. Feel free to improve the wording further. --Noleander (talk) 04:06, 30 November 2011 (UTC)
Latest comment: 12 years ago6 comments4 people in discussion

Does it specify that only major religions and languages should not be linked, or that religions and languages should never be linked? Joyson Prabhu Holla at me! 09:11, 16 December 2011 (UTC)

Not linked formulaicly ... that is, every time Christian or Jewish comes up. But clearly in List of religious leaders you'd link the religions once or twice. And probably religions that are less well-known to readers of en.WP should be linked more readily. It's sometimes a grey area. Tony (talk) 09:41, 16 December 2011 (UTC)
As with the whole of the overlink section, there is an explicit exception for "major geographic features and locations, languages, religions, and common professions" when said terms "are particularly relevant to the topic of the article". As noted, that leaves grey areas and a degree of subjectivity in respect of what constitutes relevance (as well as what constitutes a "major" religion or language anyway), but it certainly does not mean they should not be linked, "major" or otherwise.
Personally, I'd avoid linking passing references in article text where it would be hard to argue any direct relevance; perhaps with a slightly lower threshold for linking "minor" religions, on the basis that people are less likely to be familiar with them. As ever it's about context, navigability and utility. And don't forget that the articles on those religions do not simply and briefly explain what they are, which we might expect people to know already - WP is not simply a dictionary after all - but are lengthy and detailed entries that may well include useful or interesting information for someone reading the prior topic. N-HH talk/edits 17:16, 16 December 2011 (UTC)
I agree with the answers above. I think it can be expressed more simply, though: Just ask yourself: "How many of my readers will really be grateful for the additional link, as opposed to being distracted?" Rules are only needed because for many editors it's hard to put themselves into other's shoes, particularly so for topics that define identity, such as religion, nationality, and language. — Sebastian 20:55, 16 December 2011 (UTC)
Although that does rather simply move us on to the next question - how many would be enough; or, in the case of distraction, too many? I would add that I'm rather sceptical of the distraction argument anyway - how would we define distraction as opposed to interesting - and relevant - but maybe tangential connection? Are people really distracted as such by links anyway, even as far as their being somehow deceived or forced into clicking on them? And, at the lower end of that scale, what business is it of anyone else's even if they are?
To be more specific, I think an article on the Pope or Saudi Arabia that failed to link Catholicism or Islam respectively would be seriously underlinked; equally, where the article on Iran lists the various religions found there, I think it would be odd to link Zoroastrianism but not Islam, on any claimed basis that the latter was "well known" or "major". It remains a relevant and significant term - perhaps the more relevant of the two, given the numbers involved in each - and it would be inconsistent to link one but not the other. Beyond that, as acknowledged, it gets more tricky. With language, for example, even English language should surely be linked - even on the English wikipedia - in an article or article section about languages, eg Frisian languages; but not when mentioned in passing, eg "the Bible was first translated into English in the year ..". In fact, in that case, a good piped link is used (see towards the end of that section). N-HH talk/edits 15:20, 17 December 2011 (UTC)
Possibly. But always try to find a more specific link. In Saudia Arabia, for example, if there is an article on "Islam in Saudi Arabia", or a section on it, that would be preferable as a primary link. Better still, linked to Wahabi Islam (is that the spelling?), the particular brand of Islam of the Saudi establishment. It could still be piped to Islam, depending on the localised grammatical context. Tony (talk) 15:28, 17 December 2011 (UTC)
Latest comment: 12 years ago5 comments4 people in discussion

This [4] just seems wrong to me. When I click on a blue link I expect to go to an article, not an outside web page. It also defeats the purpose of red links which is to encourage article-building. Is there any specific guidance on this (I imagine there is and I've just missed it). Thanks. Dougweller (talk) 07:59, 29 December 2011 (UTC)

  • Yes, I think this is covered by WP:EL, which discourages such links in the text body. I can understand the general reticence to allow red links to exist/remain, but they are indeed useful as you said. The user clearly doesn't understand the difference between a ref and an external link masquerading as a wikilink. I remove such links on sight as 'linkspam'. --Ohconfucius ¡digame! 08:40, 29 December 2011 (UTC)
Thanks very much, and for fixing it. I'll remember that in the future. Dougweller (talk) 11:38, 29 December 2011 (UTC)
I'd keep the red links unless I thought the topic is not notable enough to ever have its own en.wiki article, in which case I'd have the external links. ― A. di M.​  14:06, 29 December 2011 (UTC)
Of course, we should not make a red link dedicated for a topic not notable enough for an article, but the correct solution depends on where and how is it short of notability. More precisely, may it be considered as a part of some other Wikipedia article – if it may, then a link to such article (blue or red), a link through redirect or a link to section are more appropriate than short descriptions and external links spread possibly by several articles. Incnis Mrsi (talk) 20:33, 2 January 2012 (UTC)

Repeat linking in reference sections

Latest comment: 12 years ago26 comments8 people in discussion

It's not unusual to see multiple linking of works/publishers in the reference section of articles. Take Fight Club for instance, which is a featured article, and there are multiple links to publishers such as Box Office Mojo and Variety in the references. It's not unsual to see this, but I recently saw a review where an editor was asked to remove repeated links from the references section. WP:OVERLINK states:

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 at the first occurrence after the lead.

At first glance, it seems to indicate that links in the reference section should only be linked the once. However, the guideline also seems to contradict itself, or at least muddle itself by also stating links may be repeated ... at the first occurrence after the lead. Now, here's the thing, because of the citation system, pretty much any citation can be the "first occurrence" of the word after the lead, as in terms of the likelihood of being encountered by the reader. Nobody apart from a reviewer is going to read through the references in order, so there is no more logic to linking a term in the first citation than there is in the last.

So I'm basically after some clarification. Are references exemopt from the overlinking guideline, or are we still limited to linking just one occurrence of each term. Fight Club has 100+ references, so if you click on a citation that takes you to a reference, is the reader really expected to go trawling through the other references looking for one with the linked term? Either way, I think this should be made explicitly clear in the guidelines. Betty Logan (talk) 20:14, 5 January 2012 (UTC)

There was a recent thread on this here [5] but with no conclusion.
My opinion is that due to two aspects: a ref list being nearly like a list or table (where we do want duplicate links to be consistent) and that ref order can change on a whim when a new reference or reused references is added, that if one is linking reference any publication/work sources (you don't have to) then all sources that aren't redlinked should be linked, even if duplicates are found. Alternatively, no ref publication or work could be linked. --MASEM (t) 20:30, 5 January 2012 (UTC)
I think the spirit of WP:OVERLINK is more geared to prose; lots of linked terms can be intrusive to the reader, but we process tables, lists and the references in a different way. The point of having references is simply so a reader can confirm a piece of information in the article, and part of that process is being able to ascertain the quality and type of source, so being able to link to publishers/newspapers etc certainly has a role. We don't really read a references section, so I don't think we should accommodate the aesthetics of the section at the expense of its function. If there is no consensus on this issue, can we at the very least stipulate that in the guideline so it is left to the discretion of editors? Betty Logan (talk) 20:52, 5 January 2012 (UTC)
I'm not sure it's possible to come to a clear conclusion based simply on the guidelines as currently written; although, as you say, generally wp:overlink is more focused on prose. Either way, I would say that it seems rather pointless and obsessive to insist on the removal of multiple links in reference notes of all places. N-HH talk/edits 11:25, 6 January 2012 (UTC)
Agreed; completely obsessive and pointless. -- cheers, Michael C. Price talk 11:56, 6 January 2012 (UTC)
The goal of WP:OVERLINK is to avoid a wall of blue text. Too much blue is annoying and distracting. Thus, if a source, such as Variety, is mentioned several times within the citations, the spirit of OVERLINK suggests that should only be blue once, in the first footnote. OVERLINK also suggests that if Variety is also linked in the prose, it is okay to also link it (once) in the Footnotes. --Noleander (talk) 14:00, 6 January 2012 (UTC)
... also, the same situation arises with links to an author's name in the References section. Let's say there are four works listed by noted author Smith. The spirit of OVERLINK suggests that only the first reference to Smith should be linked. --Noleander (talk) 14:03, 6 January 2012 (UTC)
But again, as noted by several past discussions, a wall of blue text distracts from reading prose (a completely fair, accurate statement), but references are not prose. You aren't "reading" the references, and because you can click on a footnote in prose and jump to the appropriate reference, if it is not linked but only the first instance is, you'd have to go back to find it. Thus it makes no sense to apply OVERLINK to the source or author name in the reflist: it should be consistent, either by not linking any source works, or by linking all that can be linked. --MASEM (t) 14:29, 6 January 2012 (UTC)
Agree with Masem. And if a reader is dedicated enough to check a reference (as some are) they are hardly going to be distracted by blue links in other references. -- cheers, Michael C. Price talk 14:31, 6 January 2012 (UTC)
The alleged correlation between "dedication" and "absence of distraction" cannot be proven. There is no cause and effect, and I would argue that they are wholly unrelated to each other. --Ohconfucius ¡digame! 15:19, 6 January 2012 (UTC)
The alleged correlation between the number of links and "distraction" cannot be proven either, but it seems to be regularly appealed to. It's safe to say though that people do not read footnotes as such, surely, which leads us to the same conclusion anyway - ie that it does not matter if there are multiple links; and, also, as pointed out above, they may in fact be helpful. N-HH talk/edits 16:03, 6 January 2012 (UTC)
If you can't intuit the "interference" and "distraction" due to saturation linkage in refs sections compared to a utilitarian linked scenario, you and I are clearly from different planets ;-) --Ohconfucius ¡digame! 01:36, 7 January 2012 (UTC)
Me neither. (Or, people from this planet are much more unlike one another than you believe.) Personally, I find the insistence on using archived copies and having full dates with fully spelled-out month names for publication, retrieval and archiving dates much more distracting (what's the point of having the retrieval date if you have the archival date as well, and what would be wrong with “Nov” instead of “November”, and are such pages so likely to change to make the archive thing necessary in the first place?), as well as the tiny font for quotations. As for the version with the links, the distraction of seeing a few more blue words is minor compared to the utility of instantly knowing that (say) Hiroko Tabuchi and Ben Protess are notable enough for a Wikipedia article and the convenience of being one link away to find out more about them. (BTW, why do you use bullets for single comments?)[Screw it, I'm fixing that myself] ― A. di M.​  12:35, 7 January 2012 (UTC)14:00, 7 January 2012 (UTC)
Personally, I find the insistence on using archived copies and having full dates with fully spelled-out month names for publication, retrieval and archiving dates much more distracting. I agree. but that seems to be the current convention to have all those dates and not in abbreviated form, and I would remove the retrieval dates. --Ohconfucius ¡digame! 13:51, 7 January 2012 (UTC)
No, wait: those links actually go to journalist, which I agree is crazy, but as far as I can tell nobody is advocating that, so that's a gross straw man. ― A. di M.​  12:36, 7 January 2012 (UTC)
No, wait: those links actually go to journalist, which I agree is crazy put there to illustrate how it would look where the authors were all linked. --Ohconfucius ¡digame! 13:51, 7 January 2012 (UTC)
Which is not what people are advocating. ― A. di M.​  14:00, 7 January 2012 (UTC)
Sans blague? --Ohconfucius ¡digame! 14:06, 7 January 2012 (UTC)
It's probably about time to implement the spirit of the RfC, which was to remove most of the strictures against so-called "over"linking.-- cheers, Michael C. Price talk 16:18, 6 January 2012 (UTC)
Oh, Michael, you're such a tease! ;-) New consensus??? try pulling the other one! --Ohconfucius ¡digame! 01:34, 7 January 2012 (UTC)
Read the closing admin's comments, which suggests wholesale removal. -- cheers, Michael C. Price talk 13:10, 7 January 2012 (UTC)
  • Treating ref lists as tables is not as damaging as undisciplined repeat links in running prose. I can cope with repeat links to journals and publishers in a ref list, since it's such a fragmentary layout anyway. But it should certainly not be mandatory. Tony (talk) 15:21, 7 January 2012 (UTC)
    • I think the key thing that we'd all agree on is consistency - its not mandatory article to article, but if you employ linking, then all bluelinked publishers should be linked or otherwise no publisher should be linked. You shouldn't employ partial linking. --MASEM (t) 23:57, 7 January 2012 (UTC)

Well... I propose adding “footnotes” to the last sentence of WP:REPEATLINK. Is that OK with everyone? ― A. di M.​  16:38, 7 January 2012 (UTC)

Sounds good to me. --Noleander (talk) 19:31, 7 January 2012 (UTC)
I support that; citations have more in common with tables and lists than they do with conventional prose in that they are "entry specific", so there is no reason to treat them any differently. Betty Logan (talk) 00:13, 8 January 2012 (UTC)
Latest comment: 12 years ago2 comments1 person in discussion

Why does WP:YEARLINK have a tag saying it is currently under discussion? I’m not seeing a specific discussion thread here indicating there is an active discussion on that specific topic. Greg L (talk) 05:11, 16 January 2012 (UTC)

P.S. Never mind. Back on 2 December (#Timeline). I’ve deleted the tag. Greg L (talk) 05:14, 16 January 2012 (UTC)

Irish articles

Latest comment: 12 years ago11 comments6 people in discussion

WP:IMOS guidelines are not being followed when some removing of overlinks is done. They have been mentioned above and can be found in the link at the start of this paragraph. Including the section WP:IRE-IRL. I understand that removing overlinks is important but please refer to the local MOS when doing so. Murry1975 (talk) 13:38, 19 January 2012 (UTC)

Well, no. If there's a concern that these terms will be misconstrued in readers' minds, piping is a very bad way of conveying the correct term. Readers are highly unlikely to click on a linked "Ireland" to discover that it actually means something slightly (actually majorly) different from what the pipe might suggest. It is far better to spell it out in the first place. Tony (talk) 13:56, 19 January 2012 (UTC)
Please could you explain. The article for the state is at Republic of Ireland, yet island article is at Ireland, could you explain how you would get around this?Murry1975 (talk) 14:13, 19 January 2012 (UTC)
Well, a reference to Ireland can, depending on context, of course be about either the island or the state. Given that the island article - for better or worse - currently occupies the "Ireland" space, any linked reference to the state will have to piped. I don't see why that's a problem or why we should insist on it specifically spelling out "Republic of .." in text in order to avoid such piping. Equally, I have absolutely no idea why someone tried to remove the main link to Ireland from the article on Dublin anyway. N-HH talk/edits 23:37, 19 January 2012 (UTC)
IMHO there is far too much linking to Wiki article on countries from BLP articles. For example, in the Liam Neeson article we have the text "Neeson was born June 7, 1952, in Ballymena, County Antrim, Northern Ireland, and is the.." Here we link, Ballymena, Co. Antrim and NI. Why do we need to link all these names - if the reader is really interested in finding out more about Neeson's place of birth, surely a link to Ballymena is sufficient? This is the type of overlinking I object to in a Wiki article. --BwB (talk) 03:43, 20 January 2012 (UTC)
The chained country links are one large facet of the problem, and correct me if I'm wrong, but even those that often oppose the full extent of my unlinking efforts such as Nick (NHH above) accept it as a problem. County Antrim and Northern Ireland only have secondary and tertiary relevance to the subject, and are crying out for removal. --Ohconfucius ¡digame! 04:24, 20 January 2012 (UTC)
The formulaic linking of the country-name is going to lead to misleading text; and Ireland as a nation, a culture, a people, is hard to mistake as the default meaning (if one is referring to the island itself, that would need to be marked, at least on first occurrence). In almost every case, a more specific link is required. Tony (talk) 05:49, 20 January 2012 (UTC)
I accept there's a case to be made in respect of articles about people and/or in terms of lengthy chain links (although I don't think it's necessarily clear-cut). However, neither really applies to the main case here, that of Dublin. The idea that the page needs remedial surgery for overlinking in its first sentence - "Dublin is is the capital and most populous city of Ireland" is pretty untenable. The link is there because "Ireland" is a manifestly relevant and related topic to "Dublin". The piping is necessary because the reference is to Ireland as a state, not the island. I'm not quite sure why you're suggesting/thinking that the purpose of the piped link is simply to explain what is meant by Ireland in this context Tony - which I think you might be. You see far too much through the idiosyncratic and narrow prism of seeing links as purely pedagogic rather than as also navigational (and, more brodaly, that also colours far too many of these debates). N-HH talk/edits 10:22, 20 January 2012 (UTC)
@Tony: The fact is that not everybody agrees that the culture and people of Ireland coincides with the Republic of Ireland (but not everybody agrees it coincides with the island of Ireland, either). Try telling a nationalist from NI that he's not Irish (or a unionist from NI that she's not British) and, well... in the best case, you'll find that's not a good idea. (Myself, I call them “Northern Irish” unless I know better about an individual.) BTW, I agree that the half-assed compromise whereby the sovereign state and the island are called “Ireland” and “island of Ireland” in article text but “Republic of Ireland” and “Ireland” in article titles, requiring all links (in contexts where the distinction matters) to be piped, combines the worst of both world, but that doesn't have a snowball's chance in hell of changing in the foreseeable future – it's about as likely to happen as for the six counties to become part of the Republic. :-) ― A. di M.​  10:45, 20 January 2012 (UTC)

I agree: it's weird. I don't think those guys agree among themselves about it, and it looks like a half-baked attempt one or two of them slapped up without thinking it through. They're good-faith editors, but the guideline does need some work. I've copy-edited a bit of it at the top. Tony (talk) 15:29, 20 January 2012 (UTC)

If I were you, I'd stay the hell away from there, given the years' worth of arbitration cases and other drama surrounding that issue. ― A. di M.​  18:18, 20 January 2012 (UTC)

Everything should be linked

Latest comment: 12 years ago8 comments7 people in discussion

I'm actually quite annoyed that such a concept as overlinking exists to be honest. This is an encyclopedia. I keep hearing over and over again "but nobody is going to want to click that", well, I would! I've found numerous times I am reading an article (usually an American one!) and, not being from America, I don't know where Houston, or Dallas, or other places are. When reading an article about the University of Houston, and not being from America, I should be able to get from that article to a page about Houston.

There is no real reason why there should be a limit to links. Overlinking is in my mind a pointless rule, because as an encyclopedia the goal should be to be as useful and interlinking as possible, not as consise and pretty.

Furthermore not only does "overlinking" create useful links between pages, but it also backs up the text with kind of references. If you're talking about a place in an article, then I should be able to click that place's name to verify what you're saying about that place, even if it's just to another wikipedia article.

I would also suggest that, barring common English words, a word's usefullness as a link is given by the fact it appears in the page at all - if it isn't important to the subject it wouldn't be on the page.

Don't assume that somebody "wouldn't want to visit x". It's hugely big headed to think nobody needs to know about the United States. This is an English language wiki not an American-only one.

95.175.136.29 (talk) 10:28, 26 January 2012 (UTC)

Many of us feel that way, but a small clique of determined delinkers refuse to allow any policy change here.-- cheers, Michael C. Price talk 11:05, 26 January 2012 (UTC)
Linking every single word (and I mean every, as my desktop Encarta dictionary definitions do) could work in some respects, if, like the Encarta system, linked items were not visually distinguished from their surroundings (except perhaps if one's cursor were hovered over a word). But what we'd lose is the ability to use our expertise, our familiarity with a topic and its sibling articles, to point readers to useful, relevant, and valuable targets through intelligent rationing, as we currently do. This would be a major loss of functionality. Tony (talk) 11:09, 26 January 2012 (UTC)
Links are not scarce and therefore do not require rationing. -- cheers, Michael C. Price talk 15:17, 26 January 2012 (UTC)#
This could work if there were "links" and "important links". Links would appear black, important links blue. 90.220.230.144 (talk) 17:19, 26 January 2012 (UTC)
The link structure of Wikipedia is itself precious information. A link from the present topic to another topic says that the present topic's connection to that other topic stands out in some important way from all other topics' connections to that other topic. Links wouldn't carry this information if everything were linked. —Ben Kovitz (talk) 14:38, 18 May 2012 (UTC)
See WP:PEREN#All words and phrases should be linked (though I agree that in the last few years that idea has been taken a bit too seriously and common words are unlinked no matter how relevant they are to the topic).― A. di M.​  14:09, 26 January 2012 (UTC)
There are of course two issues here - whether such a thing as overlinking exists at all; and, if so, where sensible boundaries lie. Unlike, it would seem, the OP here I accept the first premise - for example, I personally don't see the value in linking common English-language words unless very relevant to the page topic, or section of that page, at hand. Equally, while I would not bar repeating links, I think you can have excess repetition. On the first point, "A Labrador is a type of dog" is good; "while walking his dog, Einstein came up with the theory of relativity", no need, surely. The same applies for so-called "well known" countries, geographical places - eg "Guam is an island in the Pacific Ocean"; "Frankfurt is a town in Germany" etc seem sensible, logical and useful links. Fortunately, the actual wording of wp:overlink backs that interpretation, encouraging as it does "relevant" links. And wp:link as a whole stresses the role of links in binding the encyclopedia together. Unfortunately, a small number of editors have taken it upon themselves, despite what overlink actually says, to remove vast numbers of perfectly useful and valid links (including ones equivalent to those mentioned above), while citing it in support of their actions, and accusing everyone else of simply being "wrong" about the issue, rather than accepting that - shock! - people might differ in their views about what is "better" linking.
The problem seems to lie in the numerous assumptions they make - eg that certain things are "well known" to all readers; that such readers would not want to navigate to (and read) the detailed WP page on that thing even if the broad topic is familiar to them; that somehow people are duped into clicking onto links that they do not actually want to click on, and that we must remove said links to prevent that happening; that one or two self-appointed editors know better than every single potential reader of that page what other pages those readers will want to - or even should - navigate to; that people have to read a whole page and that if they only make it to one section, that's tough on them in terms of having relevant links there if it's already been linked elsewhere recently; that people can always use the search function if the thing they are interested in is not linked (sure they can - but why should they have to?). As for said editors' "expertise" and "familiarity with the topic" - the first is unsupported self-assertion; the second has no relevance of course to the familiarity that others might or might not have with the topic. N-HH talk/edits 17:23, 26 January 2012 (UTC)

City, state and country names

Latest comment: 12 years ago12 comments4 people in discussion

In my many years of dealing with music album articles, I've always linked US-based recording locations in this manner: Cotati, California; Easton, Pennsylvania; North Hollywood, Los Angeles, California; San Diego North County, California, etc. For locations outside the US, I've done it like this: Reading, Berkshire, England; Richmond, Victoria, Australia; Toronto, Ontario, Canada. Is this correct, or should the cities not be linked? I think they should, because whilst major countries like England and Canada should be obvious to native English readers, those specific cities, counties and provinces may not be. On the "Overlinking" section of this article, it states to "Avoid linking the names of major geographic features and locations", but what exactly constitutes a major location? Should, then, the province of Ontario not be linked? Again, I think it should, because not every reader will be well-versed in Canadian geography (certainly not here in the UK, where I'm sure some illiterate souls couldn't even point to Canada on an atlas!) All in all, I'm not getting much information from the quoted sentence above. It leaves too much room for ambiguity. Mac Dreamstate (talk) 15:03, 2 February 2012 (UTC)

Irregardless on the linking of the city to start, when you link the city, you don't need to link any larger states/providences/countries that it is in, as it is implicit that the city article will have these links. To be explicit, all 4 examples you lead off with are linked properly for this, but the Toronto Ontario example is wrong (it should be "Toronto, Ontario, Canada". As for what is a major city that shouldn't be linked, ask yourself if you'd expect the city to be known to anyone that is able to access Wikipedia. Also consider what importance that the location has to the actual album, since geographical location is rarely germane to the topic of the article, outside of the location of the recording studio. For your examples, I wouldn't at all link North Hollywood or Toronto, but the others are not quite as well obvious. --MASEM (t) 15:11, 2 February 2012 (UTC)
The part about relevance to the topic makes sense, and I've only recently started to consider not linking locations at all within album page infoboxes (having observed the various templates in more detail). I'm still on the fence about that, but most likely I'll add it to my list of things to start de-linking in my overlink cleanup sweeps. I try to do things by the book around here but sometimes I like to skip the pages a little bit, heh. Outside of album articles, though, are you saying that it comes down to a judgement call to link certain cities, rather than something set in (Wikipedia's) stone? Mac Dreamstate (talk) 15:27, 2 February 2012 (UTC)
Well, its basically "What are cities that you'd expect every Wikipedia to know?" I'm sure there's a short list where there's no disagreement such as New York City, London, Paris, Tokyo, etc. But there's a broader list that becomes a bit more iffy, and likely will have regional differences, like Portland (either Maine or Oregon), Manchester, Lyons, Osaka, etc. At that point I'd defer to how relevant the link is to the content. Note that if it is in an infobox or in a table as opposed to prose, it is generally okay to link even if it a large well known city , as long as that's part of the consistency for the infobox/table. The point of avoiding the links to well-known cities in prose is that they can be distracting when the term is obvious. ("Soandso traveled to New York City to record his latest album..."), but helpful if the term is unfamiliar otherwise. --MASEM (t) 17:11, 2 February 2012 (UTC)
The two key words in the guidelines are "well known" and "relevant". The limits of what is "well known" have never been defined (indeed, I'm not sure you ever could, although people seem to try pretty hard and enforce linking/delinking on the basis of what they happen to believe); and the relevance allowance overrides that anyway. In this case, I'm not sure how relevant you'd say cities/countries were to albums. Beyond that, chain-linking I guess is frowned on. But equally, the rules that attempt to cut back on "unnecessary" or "cluttering" links are obviously less of an issue in infoboxes, out of prose. And, finally, these are guidelines of course anyway, not rules of stone. N-HH talk/edits 20:51, 2 February 2012 (UTC)
So within the album infoboxes, linking to not-so-well-known locations isn't frowned upon after all? That's what I've been doing until now. Granted, there's still the issue of it not being relevant to a music article, but if it isn't necessarily such a wrong thing to do, then maybe I'll spare myself the task of going back and de-linking everything from album articles I've created/edited. Plus, I don't see why people wouldn't be interested in knowing more about a cool-sounding place like Växjö (as can be seen on the article for Heavy Machinery), heh. Mac Dreamstate (talk) 22:02, 2 February 2012 (UTC)
Links in infoboxes are generally fine even if is a large city; consistency should be sought after there. But the same idea of chain linking should be avoided (eg don't use Toronto Ontario but instead use Toronto, Ontario). --MASEM (t) 22:09, 2 February 2012 (UTC)
OK, one more thing: in the case of a more long-winded location such as Woodland Hills, Los Angeles, California, is that how it should be formatted or should California be left blank, as in Woodland Hills, Los Angeles, California? I'm not very keen on dropping the name of the state altogether because, again, for the sake of geographically challenged readers they might not even know where Los Angeles is. That, or abbreviating the state name to CA, as seems to be customary over there—not everyone will be aware of that system works. Mac Dreamstate (talk) 22:26, 2 February 2012 (UTC)
I would include the full geographic string in the link since being linked, if someone really didn't know what California was, they would easily find their way there following the first few links in the leads of the linked city articles. --MASEM (t) 23:55, 2 February 2012 (UTC)
I'd say if an English-speaker doesn't know where LA is, they should go back to infants' school. The whole string is unnecessary clutter. What's wrong with just Woodland Hills, Los Angeles? Easier to read and identify the link, yes? If there's some particular reason to link to it, this should be to a more specific section or daughter article of the city that is relevant to the topic. MOSLINK also says to avoid bunches of links. Tony (talk) 02:12, 3 February 2012 (UTC)
I'm not sure the focus on "where X is" and whether WP users might or might not know that already really helps much. WP articles on places don't simply offer a geolocation or one-line description and nothing else; and, again, it's impossible to say what individual WP users might know or not, or what they might be interested in reading more on, whether they know about the thing at a basic level already or not. Comments about "going back to infants' school" don't help much either. Anyway, when it comes to the specific questions, the guidelines aren't 100% prescriptive, and nor should they be. Equally, these responses are all just the opinions of random other editors. Some dislike chain linking that string-links to city, state/county, country; others, such as myself, aren't much fussed. None of us are experts any more than anyone else, and I'd just stick to what make sense to you, while applying a bit of common sense to the process n order to avoid both the extremes of needless cluttering with redundant links, especially in text, and of needless asceticism that limits reasonable navigability for the wide range of people who come to read WP articles, with all their diverse knowledge and interests. If that helps. N-HH talk/edits 09:53, 3 February 2012 (UTC)
We've already excluded chainlinking from this. It's one link containing three terms, linking to the most specific location. --MASEM (t) 13:16, 3 February 2012 (UTC)
Latest comment: 12 years ago3 comments3 people in discussion

Currently there are a large number of incoming links pointing at India and Turkey. For India in particular, there are a lot of cases that the articles talk about people or events of pre-1947 partition India. For Turkey, there are articles that talk about the bird. Since neither is disambiguation page, is there any way that incoming links pointing at them can be monitored? 61.18.170.228 (talk) 17:18, 17 February 2012 (UTC)

If these incoming links cannot be monitored how can they be disambiguated and modified to point to the correct page (e.g. Turkey)? 61.18.170.76 (talk) 15:16, 18 February 2012 (UTC)
They can be monitored if anyone wants to do it. You can use, for example AWB to create a list of incoming links, and compare it with another version of the same list the following day or week. This could be automated easily enough too. Rich Farmbrough, 20:43, 27 October 2012 (UTC).

Bio intros being linked/delinked to countries and constituent countries

Latest comment: 12 years ago18 comments5 people in discussion

Hello, I need clarification for my gnome edits, concerning links to bio intros. Which are proper [United States|American], [Wales|Welsh], [Canada|Canadians] for examples or [Americans|American], [Welsh people|Welsh], [Canadians|Canadian] for examples? GoodDay (talk) 08:34, 18 February 2012 (UTC)

In principle, I'd favour the "people" articles rather than the country, although a lot of them are full of ethno-nationalist rubbish and linking to them might imply that the subject is more of an ethnic flag-waver than they really are. It can also spark pointless WP arguments - eg someone obecting to a Basque painter being linked as "Spanish". Linking them as coming from "Spain" is less controversial as a statement of fact of course.
Beyond that, you'll find plenty of people on this page who think there often should not be a link there at all, especially for places like the US, on the rather odd grounds that "everyone knows what/where the US is". There are no hard rules though, and also a possible contradiction between WP:CONTEXTLINK section in wp:lead, which seems to encourage links (and did so even more before one of the regulars here went and quietly changed it); and WP:OVERLINK here, which looks to discourage them. Sorry, that's not really an answer, but I'm not sure there is one. N-HH talk/edits 11:29, 18 February 2012 (UTC)
I can't see the utility of these "people" pages. "Americans", for example, opens with, "People of the United States, also known as simply Americans or American people, are the citizens of the United States. The United States is home to people of different nationalities.". Gee. Profound. The whole point of the overlink guideline is that linking should be used carefully—rationed, as it were. There are a number of compelling reasons for doing this. GoodDay, can you provide examples of bio articles where such links are useful to readers? In addition, almost every such general link could be strengthened by making it more specific, to a daughter article or a section. I don't want to get into a debate about this with N-HH, who disagrees with the guideline on these matters. Tony (talk) 12:38, 18 February 2012 (UTC)
Well no Tony, I broadly agree with the guideline at overlink, which advises against linking things for the sake of it, but specifically and explicitly allows for links to supposedly "well known" terms (the threshold for which is never defined anyway) when relevant to the topic at hand. I just disagree with your and one or two other people's entirely subjective interpretation and implementation of it. You are well aware of, but constantly elide, that distinction. Please, finally, can you stop doing that? Here, as with all linking questions, surely the issue is one of relevance, context and navigation, not whether someone might or might not know what X is or are. Whether the links are "useful" or not is also kind of a meaningless question, because obviously any content or link should have utility; but that just begs for a definition of what we mean by that and who you define it in respect of - what is "useful" to one person is useless to another. As noted, I do agree on the point about the poor quality of the people articles. However, quoting the opening definition of one of them and pretending that is all the article says hardly goes to proves the complete pointlessness of that one, let alone of all of them; and, if they are really completely pointless, they should surely not exist at all. If the pages have any value at all, it logically follows that a link to them will be relevant in some cases. The question is when exactly. N-HH talk/edits 13:05, 18 February 2012 (UTC)
Well no, N-HH. "If the pages have any value at all, it logically follows that a link to them will be relevant in some cases." This is not logical. One or two other people ... do you have evidence for that assertion of low numbers? Please, finally, can you stop doing that? "because obviously any content or link should have utility" ... let's link every occurrence of "the", then. Tony (talk) 16:12, 18 February 2012 (UTC)
Foolishly perhaps, I will respond to your rather pithy and utterly childish - in terms of its repetition of my phrasing - comment. 1) Unless you are suggesting there are topics that stand in isolation from everything else in the world, it is fairly certain that there will be some cases where a link to the page on thing X from another page will be warranted. That's a rather unremarkable observation on my part. What is remarkable is that anyone should contest it. 2) Yes, I do have evidence. It is only you, OhConfucius and Colonies Chris and a couple of others who I see tearing through pages taking out not just admittedly repetitive, redundant or irrelevant links, but perfectly decent ones too. By contrast, I see people regularly querying that behaviour, on this page and on your talk pages; and, of course, you have persistently refused to ever point to where there was evidence of a wider consensus for this. You just assert it, while misrepresenting what the guidelines actually say, as you are doing here. 3) You were the one who insisted on links being "useful". My point about utility was agreeing with you on that. And you really let yourself down by throwing the "let's link 'the' shall we?" into the mix again. If you can't debate this seriously and can't - or won't - understand what people are saying, then don't bother. N-HH talk/edits 16:38, 18 February 2012 (UTC)

It would appear that de-linking from countries/constituent countries & people articles, would be best? GoodDay (talk) 15:30, 18 February 2012 (UTC)

Different editors will continue to disagree over how the guidelines should be interpreted. It seems to me that any changes made without a very clear and specific justification in each case may well become contentious, and are best avoided. Ghmyrtle (talk) 15:39, 18 February 2012 (UTC)
I think we can all agree that these particular pipelinks [Republic of Ireland|Irish], [Northern Ireland|Irish] & [Irish people|Irish], are certaintly best left alone. GoodDay (talk) 15:43, 18 February 2012 (UTC)
I'm not so sure about this. Ireland is very commonly known throughout the world, and especially among people who know enough English to be using this site. More importantly, piping in this way can sometimes slip up: there are historical, national, and geographical sensitivities we should not display incorrectly by mistake, by piping everything to Ireland and Irish. The first time in an article, if the meaning is Eire or the Republic of Ireland, and this matters in the context, it shouldn't be piped to Ireland or Irish, which can be ambiguous. Let's be careful. Certainly Northern Ireland is a risky pipe to Irish. Please judge on a case-by-case basis. Tony (talk) 16:08, 18 February 2012 (UTC)
It's best to skip the Irish stuff, IMHO. GoodDay (talk) 16:12, 18 February 2012 (UTC)
The way I think of these links - very common terms or names or the like but limited to the first or second sentence of the lead, typically in the immediately sentences "TOPIC is a X, Y Z" - are what I think of as taxonomic links, establishing where the specific topic falls into a large classification structure. Such links, only in those lead sentences are important to have, as long as consensus generally agrees that they apply to how we at WP organizing topics. Sometimes these will be repeated into the infobox, but not always. We should link these taxonomic terms when used like this, as to allow readers to jump to the immediately higher-up class in that heirarchial information structure. So, in the case of an Irish person, linking "Irish" to an article about people of Irish nationality makes sense, because by nationality is one way we classify people. Mind you, there's a lot of gotchas that I recognize here, but that's why its almost always what the first or second sentence of the lead where these phrases appear that linking makes sense; outside of that, the value of such links decreases dramatically. --MASEM (t) 16:40, 18 February 2012 (UTC)
Concerning constituent countries: I assume there'll be little (if any) complaints about changing links in corresponding bio articles from [England|English], [Wales|Welsh], [Scotland|Scottish] to [English people|English], [Welsh people|Welsh], [Scottish people|Scottish], however. GoodDay (talk) 16:46, 18 February 2012 (UTC)
Irregardless if they should be linked or not, yes, the nationality article should be linked, not the nation's article. --MASEM (t) 16:53, 18 February 2012 (UTC)
That's good enough for me. GoodDay (talk) 17:00, 18 February 2012 (UTC)
Ha! Wait until you get to someone from Cornwall. Are they English? Cornish? Both? And, just to clarify, since Tony seems to have taken it upon himself to completely misrepresent my views on this, in this thread and on the OP's own talk page, thus creating the very row he claimed he wanted to avoid. I am agnostic on the specific issue of nationality links in leads - whether they go to people or country articles - and genuinely don't have a view to offer beyond my original comments. And there is no firm right or wrong on the point, however expert individual editors might like to tell you they are on how to link "properly". However, on the broader point, it is flat-out wrong to suggest that either the guidelines or a separate consensus deprecates links to "well known" things full stop. There is a relevance exemption, explicitly in the guidelines. Equally, "WP is not a dictionary" actually works, if anything, against the do-not-link argument, precisely because the WP pages being linked to are not simply dictionary definitions (this point will sink in eventually for the one person who insist on repeatedly citing it). This also undercuts the "a 10 year old knows what American means" argument, which also has no relevance since the links are not necessarily there to simply define or explain what something is, but to provide context and enable navigation between pages on related or relevant topics. N-HH talk/edits 17:02, 18 February 2012 (UTC)
For info, WP:LEAD used to include a specific example of "American pianist", with a link to both US and pianist. That was taken out in a unilateral intervention a few months ago. Of the small number who then commented on the edit, the ensuing talk page discussion about this came down 4 to 2 (3 to 1 if you exclude me and Tony) against such a restricted interpretation of linking things like nationality and profession in the lead based on what every reader supposedly already knows. Not a big sample, but it's the only direct discussion on the specific question here that I've seen anywhere prior to this one. N-HH talk/edits 17:16, 18 February 2012 (UTC)
I'll just keep puttering along. Since different editors have different bio articles on their watchlists, they can always merely revert any of my changes. GoodDay (talk) 17:20, 18 February 2012 (UTC)

Linking inside titles

Latest comment: 12 years ago3 comments3 people in discussion

Looking for previous discussions or guidelines about linking inside titles, for example

The book Rose is My Wife was published in 1956

vs.

The book Rose is My Wife was published in 1956

Generally it's been my experience we don't wikilink words inside titles, rather add clarifications in parenthesis, like:

The book Rose is My Wife (about Rose McGowan) was published in 1956.

The reason being wikilinks insides titles will stop working once the title is red or blue linked; are confusing because the link breaks up the title into separate parts; and confusing to newbies who think clicking on a title should go to the book itself. Any help on previous consensus appreciated. Green Cardamom (talk) 20:03, 24 February 2012 (UTC)

I agree. It irked the hell out of me when the first link in Með suð í eyrum við spilum endalaust#Track listing went here rather than here. ― A. di M.​  20:24, 24 February 2012 (UTC)
Agreed - I can understand the need to link, but there's always a way to do it without placing the link in the title directly. --MASEM (t) 23:46, 24 February 2012 (UTC)
Latest comment: 12 years ago25 comments10 people in discussion

User:The Rambling Man believes that this Manual of Style recommends linking to "items every time in a sortable table". Now, common sense and what I understand as the spirit of this guideline seem to argue against repeating the same link to a major newspaper over 100 times, but maybe he is right. What do others think? --John (talk) 21:01, 29 March 2012 (UTC)

While this is becoming a bit forum shopping, I'm happy to contribute my perspective (for the third time for John's benefit) which is that OVERLINK allows us to be generous to our readers if useful within a table. Sortable tables enable our readers to focus on the categories of tables that they are most interested in. Now, if we link an item once and once only in a sortable table, and our reader sorts the table in a manner different to the default, the first instance of a given linkable item won't necessarily be linked. Indeed, it might be linked some way away from the only linked item in the table. Is that useful to our readers? I don't believe so. For many months (maybe years), featured lists have received community consensus with linkable items in sortable tables being linked every time. In this instance, User:John has used a semi-automated tool with an incorrect (in my opinion) rule set to remove those links. I believe his edit to be incorrect, unhelpful and therefore disagree with his approach. The Rambling Man (talk) 21:09, 29 March 2012 (UTC)
At the FLC of the article in question, I asked for these links to be added. At the time, REPEATLINK was more explicit about it than now. I'll look into when and why this changed. Goodraise 22:56, 29 March 2012 (UTC)
I would also question such repeat links. The Observer is a well-known journal, and over 100 repeats does seem rather over the top, but: 1)REPEATLINK seems to allow it, because it's in a table and 2)it's a large table. IMHO, the problem is not as irritating as if it was a table where 'USA' or 'Russia' was linked 100 times.--Ohconfucius ¡digame! 23:06, 29 March 2012 (UTC)
Okay, I've traced the change to this edit and that discussion. Goodraise 23:13, 29 March 2012 (UTC)
On a sortable table - as long as the linking of one specific column of data makes sense for all elements in that dataset (excluding the obvious cases of redlinks), then yes, repeat links should be completely fine. There is the option that Ohconfucius suggests as a consideration is that if the terms in the data column are sufficiently common/well-known, then linking may not be needed (eg. depending on scenario, a link to a country column if the article is not directly related to geography, politics, or the like); however, if it is the case that there are not-well known elements that would be linked among well-known elements, then all elements that can be linked should be linked in this table to be consistent. --MASEM (t) 00:41, 30 March 2012 (UTC)
Vertical inconsistency can look a bit messy in a table. Tony (talk) 01:47, 30 March 2012 (UTC)

My two cents I wrote the article and the reason I linked it repeatedly is because it's a sortable table, as pointed out before. It seems more helpful to me for users to be able to find out what this publication is each time its mentioned rather than having to find the single instance where it's linked. —Justin (koavf)❤T☮C☺M☯ 02:01, 30 March 2012 (UTC)

As ever, the argument that this might be useful and offers substantive functionality seems to outweigh the argument that one or two people might not like the look of multiple blues in a table, which would seem to be of marginal relevance. FWIW I think more comprehensive linking is more aesthetically attractive in tables anyway, so the appearance argument cuts both ways, depending on who you ask, which makes it even less worth bothering with. Also, as a side-point, I'm not sure the Observer is that well known out of the UK (or even within in, if you look at circulation figures). N-HH talk/edits 14:40, 30 March 2012 (UTC)
That assertion seems to presuppose that only the paper's readers know a journal. --Ohconfucius ¡digame! 15:41, 30 March 2012 (UTC)
Well they might know of it or know what it is, but I was assuming that people who don't read it don't know much about it. That may or may not always be true, but it's a reasonable guess (and, on even more of a side-point, since a lot of the more radical delinking of certain terms that goes on comes justified by the claim that "people know what X is", that's an important distinction btw, which often gets lost). N-HH talk/edits 16:47, 30 March 2012 (UTC)
OK, I'll admit it: what's the Observer? It sounds like some kind of newspaper or something. Art LaPella (talk) 20:37, 30 March 2012 (UTC)
If a genuine "observation" (irony) then fair play, but if an attempt at irony, poor (this is English Wikipedia, not British Wikipedia); why should someone (e.g. Randy) from somewhere in the midst of the US (e.g. Boise) know that The Observer is a weak Sunday pseudo-broadsheet in the UK with a circulation of around a quarter of a million? i.e. read by 1/240th of our populous? In the meantime, an Italian sports newspaper that most US citizens (the majority of our readers) will never have heard of gets a third of a million readers? What you may think is obvious is not obvious to every English-speaking reader, nor the rest of our global audience. The Rambling Man (talk) 21:21, 30 March 2012 (UTC)
No irony intended. I do in fact live 600 miles from Boise. Art LaPella (talk) 22:04, 30 March 2012 (UTC)
Aren't we getting a bit off topic here? Things like Russia, Africa, or Earth don't usually need to be linked, at all. That The Observer should be linked at least once, doesn't seem to be in question here. Whether the average reader will be familiar with The Observer is besides the point for links to it beyond the first. The question here is whether the utility of linking to it on every occurrence within a sortable table outweighs the distraction caused by the mass of blue it entails and the aesthetical displeasure it apparently causes to some of our editors. Personally, I think the distraction is much less an issue within a table than it is within a paragraph of prose, but maybe that's just me. Goodraise 21:41, 30 March 2012 (UTC)
Entirely so. The initial assertion by User:John that "we definitely do not need multiple links to the same target; one or two will suffice" is part of the origin of this debate. Of course, how to implement "one or two" links to the same Observer in this table would need to be discussed. Do you link the first and the 50th? The first and last? The 25th and 75th? I'm not sure how that would work. For me (and, it would seem, all recently promoted FLs), it's all or nothing. Sure, no need to link Earth but there seems to me to be a legitimate reason to link minority British publications in a global English-speaking encyclopaedia, and there seems a legitimate reason to do it every time in a sortable table where the first instance of the name of the publication may not appear first or near a reader's interest after a re-sort. Of course, for prose this is an entirely different discussion. The Rambling Man (talk) 21:57, 30 March 2012 (UTC)
It seems to me that the two arguments for linking every instance of a linkable term in sortable tables are aesthetic ("Vertical inconsistency can look a bit messy in a table") and utilitarian (Randy from Boise will not know what The Observer is, will not properly appreciate an article about Orwell's writings without being able to click to our article on the paper, and will not be able to find this link unless every single instance of the term is linked, because he may have resorted the table). The Rambling Man says it should be "all or nothing". I disagree; I am sure that Randy will be able to find one link somewhere in the table if it's so important to him. This is what I understand the guideline currently to recommend. The spirit of the guideline is that for the majority of our readers who are not Randy from Boise, the multiplicity of low-value links will distract from the fewer high-value ones, and I am still struggling to see a cogent argument to depart from this in sortable tables, any more than in prose. When I read prose I often do not start from the top; for example on biographies I often start at the end and work backwards. In doing this I would not try to enforce the practice of linking every instance of a link, just for my convenience in case I see an unlinked instance of The Observer before I see the link; why should this be any different? --John (talk) 23:26, 30 March 2012 (UTC)
So your suggestion on how/when to link is what exactly? Every fifth instance? Every ninth? What's your practical solution rather than your individual interpretation of the spirit of the guideline? The Rambling Man (talk) 09:37, 31 March 2012 (UTC)
I'd go with once, as I said above. --John (talk) 10:15, 31 March 2012 (UTC)
So the point of OVERLINK as a guide and an aid to our readers is what? The Rambling Man (talk) 10:28, 31 March 2012 (UTC)
the multiplicity of low-value links will distract from the fewer high-value ones, and I am still struggling to see a cogent argument to depart from this in sortable tables, any more than in prose. What's distracting for someone in prose might not be distracting for them in a table, and vice versa. (Plus, what's distracting for someone might not be distracting for someone else – people are not all alike.) ― A. di M.​  17:03, 31 March 2012 (UTC)
  • I agree with N-HH here: to expand, aesthetically, either all blue (or none) in a table column seems to be better that the patchy effect. The sortability of a column can make it worse when inconsistently linked. A table doesn't present the problems of reading difficulty and dilution that we have from linking in running prose (here, N-HH doesn't agree, I know, and we'll have to accept these different positions). Is the item linked anywhere else in the text? If so, it may not be necessary to link every instance in the table. Tony (talk) 11:18, 31 March 2012 (UTC)
    • That's a great suggestion, Tony. It is the low-value nature of the link as much as the 100+ repetitions that offends this guideline. In a case like this it might suffice to have a sentence outside the table that says "Orwell published over 100 articles in The Observer" rather than linking every instance in the table. The guideline currently says "links may be repeated in ...tables..." (my emphasis) and not must. Certainly it doesn't recommend linking every occurrence. Do those who hold this view believe that we should insert something along the lines of "In sortable tables, every instance of every linkable word must be linked"? --John (talk) 11:37, 31 March 2012 (UTC)
      • No, The Observer wasn't linked beforehand, and per many of our readers who don't know what The Observer or The Adelphi is, these are very much not low-value in nature. The Rambling Man (talk) 13:20, 31 March 2012 (UTC)
      • There is the option that, say, a column in a long table only derives values from a set of 3-5 elements, of having a catch-all statement right before the table to explain that column and provide singular links there, avoiding linking in the table. (Any more than 5, and you've created a sea of blue in prose, so that's not a good option). As Tony says, internal consistency is the key guidance here, since we're not talking about well-known geographic terms. If the page editors feel The Observer needs linking in the table, that's fine, as long as all other items within the same column that can be linked are linked. --MASEM (t) 14:08, 31 March 2012 (UTC)

Appropriate way to mark dead links?

Latest comment: 12 years ago3 comments2 people in discussion

I ran across a dead link here (the last reference). I've seen a few links marked as dead but before I did so I thought I'd look around for the proper way of doing so. I've not be able to find anything, but this article looks like a natural place for that information. How should one mark dead links, and can this information be added to the maintenance section of this article? Garamond Lethe (talk) 22:13, 31 March 2012 (UTC)

Template:Dead link. ― A. di M.​  22:34, 31 March 2012 (UTC)
That's what I was looking for. Thanks! Garamond Lethe (talk) 23:39, 31 March 2012 (UTC)

Underscores

Latest comment: 12 years ago3 comments2 people in discussion

This seems never to have been discussed. Some editors have the habit of separating words in a link, not with spaces but with underscores. They probably do this because they're familiar with some software that requires underscores. But here on WP, they're not necessary. Plus, it looks terrible when the link isn't piped (e.g. President_of_the_United_States), and for those who spend any time attending to aesthetics, it means extra work for them to remove the damn underscores.

Can MOS mention this matter, please? My preference would be to BAN them. That's probably unrealistic, but can we at least say they're undesirable, unnecessary and unattractive? -- Jack of Oz [your turn] 22:05, 3 April 2012 (UTC)

A Bot like UsefulPixieBot could probably remove these on sight. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:53, 11 April 2012 (UTC)
Maybe, but that is not related to what I'm asking. Underscoring is surely a stylistic issue, and a MOS ought to have a position on it. -- ♬ Jack of Oz ♬ [your turn] 20:54, 11 April 2012 (UTC)
Latest comment: 12 years ago15 comments7 people in discussion

I propose modifying WP:OVERLINK to make an exception to:

Avoid linking the names of major geographic features and locations, languages, religions, and common professions

for the first instances of such terms, in infoboxes. For example:

Birth place = Birmingham, England
Death place = Derby, England

Thoughts? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:51, 11 April 2012 (UTC)

Well, first, you'd never link the first one like you did. You don't need run-on blue links for a city and its country, but instead simply use Birmingham, England for a single link, and the same for the second, since it's expected that the country would be linked in the city article. That said, if the place was still well known, like New York City, then linking the first time in an infobox would be reasonable. --MASEM (t) 13:20, 11 April 2012 (UTC)
First: on the contrary, I would, and I recommend that we codify that in the MoS. Your [[Birmingham|Birmingham, England]] example is particularly unhelpful and I would never do that; indeed, I'd remove such when I saw them, which is thankfully rare to the point of extinction. It's even more important to link less-well known places than the well known; but who's to say what a reader in other countries knows or does not know? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:40, 11 April 2012 (UTC)
I think it's fine as is. We don't need to link to obvious things. It's just distracting. Even the Birmingham article doesn't link to England. There's always the search box for the person who wants to know what England is Bhny (talk) 13:56, 11 April 2012 (UTC)
But, as I always have to say when this comes up, linking isn't just about explaining what something is to people - it's about navigability to related/relevant topics, as the guidelines expressly allow for, even for "well known" terms. Plus of course, WP articles don't just provide dictionary-style definitions, they're detailed and lengthy profiles. I know what England is and am relatively well educated when it comes to history, but can guarantee you half of what is on the page would be news to me. It doesn't simply say "country in northern Europe, part of UK". Also, how do we decide where the bar falls? Do we have an ideal, average reader who we assume knows what England is but maybe not Belgium? What if 20% of people do know Belgium (and how would establish that anyway) - is that enough to not link Belgium? 50%? People come to WP from 1001 different backgrounds, including 10 year-olds from Papua New Guinea no doubt on occasion.
As for infoboxes specifically, I'm personally easy about more frequent linking there, to the point where they act as navboxes almost. Plus of course, having links there can justify culling the perhaps weaker related links from the main text, which should make it easier to come to agreement on that. N-HH talk/edits 14:16, 11 April 2012 (UTC)
ps: the fact that England is not linked in the main text of the Birmingham page btw strikes me as an omission. N-HH talk/edits 14:19, 11 April 2012 (UTC)
The point that I think has been determined from past consensus (I think, I'd have to search to double check) is that we avoid back-to-back links of the style [[Town]], [[Country]] in favor of [[Town|Town, Country]] when the reason for inclusion of the link to the town is germane, such as here indicating the place of birth or death. This is not trivializing the country link, but the fact that the country like is far less germane in most cases when the town is given immediately before as in this example, and there should be reasonable expectation that the country the town is in will be linked in the town's article. --MASEM (t) 15:10, 11 April 2012 (UTC)
The problem isn't necessarily that readers don't know where England is, or what's in the England article. As pointed out by Masem, this dreadful habit of carpet-linking of geographical topics to successively less germane articles is linking for its own sake. Navigational links are well and good, but links really ought to be germane (i.e. relevant in the first degree); the tendency of using wikilinks as some sort of dictionary function is not what we are about. --Ohconfucius ¡digame! 15:30, 11 April 2012 (UTC)
"There's always the search box for the person who wants to know what England is " - Some should let Sir Tim Berners-Lee know how he's been barking up the wrong tree all these years. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:38, 11 April 2012 (UTC)
As I thought I had pointed out (above and separately ad nauseam previously; and as any cursory reading of any half-decent WP page would make obvious) Ohconfucius, the point is precisely that WP links do not simply provide dictionary definitions. The - entirely correct - "WP is not a dictionary" argument actually reinforces the point that links can be useful above and beyond simply explaining what something is in basic terms. It doesn't count against providing links where they are to terms that are sufficiently relevant - or even, more tightly, germane - and I still have absolutely no idea why people cite it as an argument against linking. "The golden retriever is a type of dog" should link to dog, "Marx met Engels while taking his dog for a walk" probably shouldn't. But the reason why in each case has nothing to do with whether people know what a dog is or not, or whether the dog page simply says "yapping thing with four legs". And I agree with Andy 100% about the old "people can use the search box" argument. Sure they can. But why make them when we don't have to? The internet has functions here that can help people that paper encyclopedias do not have. Why on earth would we be making things more difficult? And then, truly bizarrely, justifying it all by proclaiming that it was supposedly more useful for the reader (as we-who-know-best define that)? N-HH talk/edits 21:19, 11 April 2012 (UTC)
I'm going to step out of this one. I didn't realize how often it has been discussed. Others are more qualified here Bhny (talk) 22:26, 11 April 2012 (UTC)
Nick, we'll never achieve a confluence, I fear, because of a basic difference in angle about the reading experience. There are several disadvantages of linking that need to be weighed, in every instance, with the advantages. Tony (talk) 05:21, 12 April 2012 (UTC)
Well in fact of course one thing that gets obscured sometimes is that you and I do both agree on the benefits of cutting out the many redundant and repetitive links that end up getting plonked all over many articles - the difference lies in where we'd put the bar and what justifications we'd use for losing or retaining (eg I'm opposed to blanket removal of links to geographic or other so-called "common" terms - where they are proper nouns rather than ordinary English words - especially from articles where they are clearly relevant and directly related to the main topic, as are the guidelines themselves). More on point, as noted above, I'm certainly open to having more links in infoboxes, not least since this makes it easier to remove those same links from within the main prose text of articles, especially the lead, when they're of 50-50 significance or relevance. N-HH talk/edits 11:38, 12 April 2012 (UTC)
ps:@Bhny I don't see anyone really as being more qualified than anyone else. Some of us might be better able to bore on the topic, but we all have valid views as readers and editors on how to make this place work better - indeed, as an occasional editor, I came to the wider topic of linking as a reader first, a little confused as to why people were sweeping through articles removing all links to France even from articles specifically about French geography and/or things from France; and then being a little confused when I was told that some editors knew how to link better and that I should simply trust them on that. N-HH talk/edits 11:38, 12 April 2012 (UTC)
I would much prefer to see a guideline that helps a reader weigh the advantages and disadvantages, than a blanket rule. —Ben Kovitz (talk) 08:54, 19 May 2012 (UTC)

Easter egg guidance needed

Latest comment: 12 years ago4 comments3 people in discussion

I made this edit to avoid what I thought was a deceptive easter egg, but it was reverted by an administrator. The current piped link appears to go against the advice under Intuitiveness in the Piped links section and against Wikipedia:Writing better articles#Principle of least astonishment - but maybe the consensus has changed, or maybe Plot sections can be more relaxed? Could those more familiar with this issue please advise and possibly suggest clarifications in the MOS, as I still believe I made a reasonable edit, and that the current state still has BLP issues. -84user (talk) 19:24, 14 April 2012 (UTC)

That link is useless and shouldn't be there anyway. It makes you watch an ad then plays an unrevealing clip. I deleted it Bhny (talk) 20:48, 14 April 2012 (UTC)
I think they were talking about the [[Julian Assange|a blonde-haired rat]] internal link, not the external link. I agree such confusing piped link are Evil. ― A. di M.​  15:11, 16 April 2012 (UTC)
yes sorry, I missed that and you are correct. that piped link is confusing and should not be there Bhny (talk) 15:43, 16 April 2012 (UTC)

Names within names

Latest comment: 12 years ago2 comments2 people in discussion

I propose an addition to the section "What generally should not be linked": names within names. If there is a building, street, airport, park, other geographical entity, tool, or pretty much anything named after someone or something else, do not link to the name within the name. For example, all of the following links should be avoided: Hancock Airport; Wilshire Boulevard; Maimonides Medical Center; Mount Edith Cavell; Hoover High School; Mahatma Gandhi University; Humphrey–Hawkins Full Employment Act; Esquipulas Peace Agreement; Rio Conference; I Love Lucy; The Lion in Winter; The Pink Panther; Cascade Elementary School; Phillips screwdriver; Ferris wheel; Teddy bear. Whether the outer name would be a red link or blue link, the link should be to the outer name or not at all. I look forward to a constructive discussion here to generate consensus and refine the wording.—Anomalocaris (talk) 02:32, 19 April 2012 (UTC)

I would think this follows from a previous section about linking a specific word in a title of a work. There's nearly always a better way to organizing information to avoid this type of link - if the inner name is germaine to the article, explain that reasoning; if not, you can usually assume that the linked inner name will be on the page of the outer link. --MASEM (t) 04:02, 19 April 2012 (UTC)

Linking sense

Latest comment: 12 years ago52 comments9 people in discussion

This talk page and the Linking guidelines could drive one nuts. I recommend that we drop most of the strictures on formally functional linking. Leave it to good sense and the editor (and I don't mean cruising red-pencillers, but primary editors), except where there is some gross SNAFU such as self-reference.

  • The objection that blue links break the flow of reading is valid only until the reader learns what the blue means. Pretty soon he doesn't even see blue until he thinks: "Uh. What the blip does that mean? Oh, there's a link; goody!"
  • Granted gross overlinking is stupid and sets teeth on edge, but it is rare and otherwise essentially harmless, venial.
  • Even slight underlinking is worse than gross overlinking and can be harmful and confusing for a number of reasons. Certainly at the very least, by the time one has scratched one's head and decided to type in a speculative search for the meaning of the term to see whether the text means what one thinks, the disruption is far worse than skipping over a few blue or even red links.
  • Finding the intended sense of the reference where a link is omitted, is not as simple a matter as finding a word for which typing in a search finds an article. It might take several tries with different wording from what is in the text that had puzzled you. And then your search might have retrieved something that for obscure reasons is misleading. A properly formulated link would take you directly to the best (and in particular to the intended) article, often just by hovering over the link, if the lede is any good.
  • I propose, much as foregoing texts have done, that for key words in long articles at least, there (usually) be one visible link per section (including lede) unless the sections are ridiculously short; and usually one link per visible page for important links at least. For one thing, suppose I have in fact remembered that there was a link on a previous page (not having skipped down to the fifth section) and I suddenly see this linked word (which I know fairly well) but I think: "Hm? What can he be thinking?" then being able to click or hover can often bring quick, smooth enlightenment.
  • Red links certainly are disruptive, but they damwell should be disruptive. They point out work to be done and warn the reader of a gap in the material. Omit them and the work remains undone longer and more often.
  • The fact that a link refers to a major item of common knowledge does not necessarily disqualify it. If I link to Latin in a trivial sense in passing, such as "many classical writers in Latin, Greek, Sanscrit, but not often in Arabic, used this form..." it is OK not to link (though not harmful as a rule), but if I am referring to a technical point concerning the link, or it might not be clear whether the reference is to one usage of the word or another, or the usage is not generally familiar, or if in doubt, then a link is certainly justified, and for Senor Redpencil to say: "Ah, Latin! I know what Latin is; I'll unlink it!" is not only arrogant, but destructive. And if the link is to Diptera, an entomologist saying: "Hell, everybody knows what the Diptera are, and the Nematocera, and the Nemestrinidae; I'll remove the links and avoid offending anybody or breaking the flow. Everybody knows where Cape Town is, right? And yet every second travel agent has fielded requests for tickets to Cape Town where the intention was Cape Cod (or vice versa). The "common knowledge" criterion should only be used with deep, deep reserve.
  • I suggest that removal of links be tolerated only where they reflect obvious major error or nonsense. To hell with fussy concerns about redundancy.
  • As for adding links to unfamiliar articles, where the original author omitted them, that is generally OK; the mere fact that it occurs to me that it might be helpful or necessary means that someone else is likely to need it. But it can be tricky if one does not realise that it was not the link appropriate to the intended sense; when in doubt when writing, link, and make sure that you have chosen the link that is appropriate to what you are writing.
  • Another thing: using links for refs. Sure we cannot link to a WP article's text for WP's verifiability, but a lot of cruising editors simply don't understand that a link instead of a reference can be perfectly valid. Suppose I am writing about Hibernation. I deal with a point on thermodynamics. This is not the place to explain thermo, so I link to the thermo article. I don't have to include twenty of the refs in the thermo article in my hibernation article for verifiability; the thermo article should have its own refs, and for me to duplicate them (even if I am competent to do so) simply creates a major maintenance problem.

Probably I am overlooking half the important cases, but I am rapidly becoming less and less tolerant of link removal, red or otherwise. JonRichfield (talk) 09:19, 24 April 2012 (UTC)

I do not agree with "red-pencillers – primary editors" distinction. There are many users which create content in (some relatively few) articles, and act as "red-pencillers" in many other articles. The primary content just arrives in different qualities: somebody makes it MoS-compliant, but less MoS-aware user leaves a lot of room for red pencils. The task of MoS-es is to make more MoS-aware users, why is it not good?
The second point: how do you understand the problem of unlinking? Certainly, if a user unlinks some places as his primary task, without real improvements in articles, then it is bad. But in the course of an article's improvement a good editor can and should use his/her discretion about words which deserve a link, and words which do not. More precisely, there are many words and other possible sockets for a link. Some are more link-worthy, some are less, some may not have a link at all (such as punctuation marks). But each link marked by the browser distracts an attention. So, if one would link all link-worthy places in a certain article, then the article becomes less navigable (that means a reader sees too many links and makes his choice for the next jump difficultly). So, when I improve an article, I usually add links to more link-worthy items, but simultaneously remove links from less link-worthy places. Incnis Mrsi (talk) 20:24, 24 April 2012 (UTC)
Incnis Mrsi, you're not the problem, as what you describe involves approaching the concept of linking with common sense and appropriate thought and discretion. The problem lies in the (very small) group that is rigidly adapting and applying the linking guideline to endorse their methodically marching through the project using scripts to strip away links "just because" they don't see them as necessary. It is a troublesome mentality, one that has seen (as a recent example) the article on North America losing links to "continent", "South America", "Europe" and "Asia" among others. --Ckatzchatspy 21:43, 24 April 2012 (UTC)
While also leaving bizarre inconsistencies, such as leaving one language/country linked, the next not, on the basis of some arbitrary bar of what the average reader supposedly knows. Not that even that is a relevant consideration - as noted 1000 times, if relevant to the topic, as the guidelines explicitly say, such links should be there, regardless of how "common" or "well known" the term is. The purpose of linking is not simply to explain to some putative ideal reader of average education or intelligence what a word might mean (WP is indeed "not a dictionary"; and how the hell would you set that average standard anyway, even if that were the point?) or even necessarily just to provide context to the current page; but to enable and facilitate navigation to other encyclopedia articles on related topics. Still waiting as well, after two whole years roughly, for a pointer to where consensus was established for people to run these scripts or otherwise perform such extensive blanket delinking of the same terms across thousands of articles, regardless of context. As noted, common sense, editorial discretion and appreciation of context are the key to decluttering "overlinked" articles that need it; as well as a less rigid and formulaic idea of what linking is supposedly for within an online reference work in the first place. Not everyone is going to agree on exactly how to apply those principles in each case, but that's got to be the starting point, rather than "everybody knows what Europe/the English language/a dog is, hence I'm going to run a script to remove links to them or take them out automatically whenever I see them" - even, for example, from the articles about Italy, the Welsh language and labradors respectively. N-HH talk/edits 07:22, 25 April 2012 (UTC)

So far we are all four on the same side. I agree with all three. My major reservation is that one needs a positive reason (reason as in "reasoning", not just as in "excuse")to remove a link, whereas a reasonable doubt is an adequate justification for including or retaining a link. N-HH's example is very good. As you all demonstrate and urge, functional reasoning in either case is more important than passively following a formula. This principle applies to a lot more editing activities than linking, but the latter are a sorer point with me than some others at the moment. It would be nice if the article could emphasise that concept more strongly. Partly because I am not one of the policy-setting community and partly because I am too partisan to be dispassionate, I have recused myself from going further in than the talk page. JonRichfield (talk) 08:37, 25 April 2012 (UTC)

I don't see four, actually. I see you and the two regulars here who start a wave of complaint about every three months. It's always the same theme. But there's been a strong trend on en.WP over the past five years to take much more care in wikilinking. This has been highly successful in making the linking system work much better for our readers. I couldn't have put it better myself than Incnis Mrsi, above: "So, when I improve an article, I usually add links to more link-worthy items, but simultaneously remove links from less link-worthy places." It is widely accepted. Tony (talk) 09:09, 25 April 2012 (UTC)
Tony, you know there's far more than "two regulars" who have issues with this; and surely you noted that neither I nor Ckatz started this thread. And yes, I may complain occasionally - or back up the new faces who regularly come along to query what's going on - but I don't bulldoze my way through thousands of pages relinking things I'd prefer to see linked in the same way that you and one or two others do for delinking. I have the humility to accept that I have an opinion that not everyone might agree with - I do not believe that I am necessarily right and that my preferences should be imposed on everyone else. The real minority here in terms of numbers, and the one that really matters in terms of consequences, is the one that has set itself up as a quasi-official "expert" panel and makes substantive changes across the encyclopedia on its own say-so (as opposed to merely commenting on talk pages occasionally). You still haven't pointed to any consensus for what you and one or two others do, despite still claiming, even now, that it is "widely accepted"; plus I'll add to that ongoing request with a new one - any evidence for the rather bold assertion that what you do "has been highly successful in making the linking system work much better for our readers"? Not just a theoretical argument/claim about how reduced navigability is supposedly somehow more useful (which would be interesting enough), but real, hard evidence, assuming that's even possible? For once? N-HH talk/edits 11:37, 25 April 2012 (UTC)
Yes, it would be terribly useful to survey readers (not just editors) about how they use links, whether they'd prefer more or fewer of them, and so on. Otherwise all general discussions about linking are based on speculation and generalizing from one example i.e. ‘I like it this way therefore so do readers’. ― A. di M.​  11:54, 25 April 2012 (UTC)
Hi A di M. You have a good point and thanks for the reference to a good (and generally sound) read. It certainly is reasonable to consider the reader, or at least one of the readers. Or a lot more than one of them. Or as many as one usefully can. However, how cogent would a survey be? Which fraction of the audience should one write for? The largest? The cleverest? The least well-informed? All I can do is write to make the most of such quality, information and reason as I have at my command. As frequently happens, someone later materialises and improves on my efforts, and bless that one, say I. Occasionally someone makes a mess of the work and I revert or raise Cain. And sometimes there is a long-standing shouting match or a briefer discussion that settles down amicably. In writing the next article I still try to achieve a well-reasoned, readable, helpfully connected article, ignoring people who say it is too hard unless they follow the links and those who say it is too puerile because they don't think the links are necessary. By all means let anyone adjust this aspect of WP on the basis of surveys, but I am left in doubt about how we should predicate our work-a-day activities on the outcomes. Hamilton, Maxwell, Milton, Boltzmann, Gauss... I cannot see many of them radically adjusting their products in the light of the polled transient opinions of their public. Galileo maybe, but see what good that did him! But I am listening in case you have compelling counter-examples!  ;-) JonRichfield (talk) 12:53, 25 April 2012 (UTC)
The main job of Hamilton, Maxwell, Milton, Boltzmann, and Gauss was not communicating with the general public; that of Wikipedia is. See the tooltip text of http://xkcd.com/1028/. ― A. di M.​  13:46, 25 April 2012 (UTC)
Hi Tony, nice to meet you, I am sure. I don't know whether you are requesting help with your enumeration, but until you arrived there were just four of us (3+1 if you like) and I cannot include you as a fifth as yet. I am afraid you will have to earn your place. For a start please explain whether you disagree with any substantial point made so far. If you do not, then once your counting is up to scratch, welcome to the club. If you do, then please explain why you think that the current situation is tolerable in such a point of disagreement, let alone why you think there is a strong trend to take much more care in linking and making the system work much better etc. Most of what I have seen has been removal of redlinks, removal of "obvious" links on a basis of arbitrary opinion, and removal of "redundant" links because the same link occurred earlier in the article... like about two pages back in a different context. Not much for regulars to base quarterly complaints on, I grant, especially if they keep on complaining about the same things. How does it come that they do that, I wonder? Surely if the system had been amended or if the error of their views had been demonstrated, they would have stopped by now? Speaking strictly as a newbie among seasoned authorities I am of course limited to my own experience of the currently sophisticated and improved system, and commensurately constrained in expression of my opinions -- I accordingly beg your patience; could you please explain how you would correct my views on points about which I had written, and remarks as made by the other errr... two... three? A cogent word from you now might save us all a lot of time and repetitive complaint, not to mention leading to major improvements any articles that we respectively work on. Thank you in anticipation. JonRichfield (talk) 12:17, 25 April 2012 (UTC)
Delinking via scripts should never be performed, short of removing duplicate links in an article. It's one thing if it were dates where it is obvious what the content of the link is, but a computer script will never be able to judge context and value of word-based links. If there are editors using scripts in this manner, this behavior is highly frowned upon by both the community and ArbCom (see BetaCommand cases for such examples). (This is to make no judgement on the rest of the arguments, only the automatic nature of removal) --MASEM (t) 13:25, 25 April 2012 (UTC)
Now that I could resonate with! Mind you, I am not sure I accept simply removing duplicate links in an article either, automatically or otherwise; some of the foregoing examples explain why. Duplicate links in a section... maybe; or perhaps flag them. Duplicate links on a page... again, maybe. Raising a flag if a word is linked ten times in an article, maybe. And maybe never anything beyond raising a flag. But short of such examples there might be reasons for deliberate linking, and as long as there is room for disagreement, I don't see how any script can cut it. JonRichfield (talk) 13:50, 25 April 2012 (UTC)
Oh Masem, stop twisting things. Scripts involve human oversight, preferably skilled scrutiny of an automatically displayed diff—in that respect, they are very different from switching a bot on and going to bed, letting it run without oversight (that is what, I've heard, goes on ... I run a Mac, so have no knowledge of them). I'm surprised that you're making claims like this.

I see in JohRichfield that we have a link maximiser—someone, like N-HH and CKatz—who'd be happy for wikilinking to be boosted back to levels resembling the old days, when it was virtually "link what you like", without discipline. Ain't gonna happen. Linking is now recognised, thankfully, as something requiring skill, and the community has decided that linking needs to be rationed if it's to function well for readers. Linking "France" or "United States" at the top of every BLP on a French or American person is just the kind of loose approach that makes readers wonder what "that blue gunk" is. Dilution through careless linking for the sake of it is now frowned upon; and country-name links, for example, have often been lazily inserted in disregard of a more specific link.

I have a limited time-budget for this next wave of lobbying by the two main players, plus you. Someone else is missing, but I can't remember who. It's just not productive. We all want the same thing: an optimal linking system; we should be singing from the same song-sheet, not taking adversarial positions. Now I've taken time to explain some of the context to you because you're new. The others know this merry-go-round well. I'm not prepared to invest hours of my time yet again, unless there's suddenly something new on offer. Tony (talk) 15:04, 25 April 2012 (UTC)

Even in BetaCommand's case, where he claimed that he oversaw the changes made, the use of automated scripts to do cleanup tasks like this was frowned on to the point that he's been blocked from the project for a year. Yes, there was a second contributing factor, that being the inability to communicate well with users that disagreed with him; that's not 100% the case here but there's enough similarities between the two that the path between what we have now and what the situation was in Beta's case is there, we just need to avoid it. Some of the arguments presented, where on a page about a geographic feature where links to other geographic features are being removed via script, call into question whether these scripts are being closely monitored or not. --MASEM (t) 15:13, 25 April 2012 (UTC)
Tony, with regard to your comments above:

"I see in JohRichfield that we have a link maximiser—someone, like N-HH and CKatz—who'd be happy for wikilinking to be boosted back to levels resembling the old days, when it was virtually "link what you like", without discipline."

This approach - repeatedly making unsupported claims that have no connection to the poster's actual position, but which are designed to distract attention from the topic at hand - should be avoided in the interest of rational debate.

"Ain't gonna happen."

I've yet to see any consensus declaring you, Ohconfucius etc. as the ones who get to decide this. I was under the impression that this was a collaborative effort.

"Linking is now recognised, thankfully, as something requiring skill"

No-one's questioning that there is skill required for successful linking. The debate is actually about whether the arbitrary practice of wide-scale elimination of links you don't like is really supported by the community as a whole. I would like to hear some explanation from you as to how you can possibly justify the removal - for example - of links to continent and South America in the article about the continent of North America. It is clear that the scripts are not being applied with due care and attention, as evidenced by the rate at which articles are processed and the number of inappropriate removals that occur as a result of this haste. --Ckatzchatspy 16:37, 25 April 2012 (UTC)
“Scripts involve human oversight”... then you must be a damn fast reader, as I've seen you use scripts to edit a dozen articles within a couple minutes. <gd&r> ― A. di M.​  17:12, 25 April 2012 (UTC)
Relinking in an article (post lead, prose only and ignoring tables/lists/captions) is actually a bad practice. Through other parts of our MOS, we write pages to be read through as a single article at one time. For example, if discussion a notable building and the architect is named, we refer to the architect once by their full name, and then all subsequent references to that name are via a last name basis, even if the name is not rementioned until several virtual pages down the article body. For the same matter, relinking should be avoided in the same manner, simply because we start from the assumption the reader is reading through the article from top to bottom. Maybe if the article had some weird structure where something about A was talked about first, then completely disparate subtopic B, and then a return to A, may make sense to reintroduce links from the first A section into the second, but in such cases, I bet it is better to re-organize the article to put both A sections together and move B elsewhere. It's not a hard-fast rule, but it is one that the relinking of a term within prose is the rare exception, not the norm, to the point that bots can with that (though like all good bots, there should be opt-out features for editors). --MASEM (t) 15:23, 25 April 2012 (UTC)
When we pin people down, I think there is a consensus that North America should be linked in the South America article. Most links to articles like United States aren't like that; they are more likely to be people or events mentioned only in one paragraph or sentence who need to be identified as American. When we pin people down, I think there is a consensus that those links, which are the great majority of United States links, should be delinked. And the only practical way to approach that goal is with AWB, although some AWB users aren't previewing enough. I hope this reduces some of the arguments repeating for the hundredth time. Art LaPella (talk) 20:20, 25 April 2012 (UTC)
It is not specifically the tool as long as it requires human interaction to validate each change. An unattended bot that removes any link to continents, for example, would be bad. Using AWB or the equivalent itself isn't, but referring back to the Beta case and what's being reported here, when such changes are done in rapid fire manner, it suggests that the user is not review each change appropriately prior to making it. There are places where linking to "North America" on pages not about geography might be appropriate, for example.
Proper linking is a skill of art. It is not easily done. Because of that, short of patent nonsense/easter egg links, chain linking, and other easy-to-spot-and-remove link problems, linking trimming/adding is the type of thing best done under a consensus based approach such as when taking an article to GA, FA, or a peer review. An article that hasn't hit these points should be allowed to develop without trying to stymy its links at the early stage. --MASEM (t) 21:00, 25 April 2012 (UTC)
One out of 1127 articles is featured, and one out of 274 is "good". For the rest of Wikipedia, "easy-to-spot-and-remove link problems" are the overwhelming majority in real-world articles, though seldom mentioned in debates. Art LaPella (talk) 23:24, 25 April 2012 (UTC)
There is no deadline. Outside of the truly problematic link issues (repeated linking, easter eggs, serial links, etc.) that can significantly interfere with comprehension, there is no rush to add or remove links from articles. Yes, by all means, let's encourage editors to engage in smart, germane linking by providing sound advice when to add and when to remove such links, when they are creating and editing articles. But because the refinement of what is the best possible link is something that can only be done by multitude of editors working together, we might as well wait until we reach these points of quality control to actually evaluate the appropriateness and density of links. --MASEM (t) 23:30, 25 April 2012 (UTC)
There is indeed no deadline, because Wikipedia grows much faster than anything like featured articles. We will never feature everything, nor will we ever submit every word to a committee faster than new words can be written. Once again, some AWB users aren't previewing enough (based on what I've read here), but we shouldn't let the perfect be the enemy of the good. In most cases, the United States should be delinked, and overly trigger-happy AWB users should be dealt with as individuals rather than essentially give up on nearly all of our articles. Overlinking can be fixed in the future, but so can underlinking. Art LaPella (talk) 23:39, 25 April 2012 (UTC)
Again, it is not the tools that are the problem here, it is editor that use them with the appearance of automated behavior to add or remove links that are otherwise not immediately problematic and/or following a personal set of linking agenda that they know is contested. That is, using AWB in a rapid fire manner to remove easter-egg links: good. Using AWB in a rapid-fire manner to remove continent and country names from articles: not good. Until we have a casebook that describes specific consensus-agreed upon removals (and not just based on the concept of germane linking, but highly concrete guidance), no one outside of regular editors of the page in question should be manipulating links in that manner. As long as the page is still readable, there is no harm. That's why I do agree with removal of chain linking, easter egg links, and other known problems for comprehension, but when there's been a good faith effort for other types of links, one editor using a script-enabled tool should not be the judge jury and executioner of that, particularly in such a rapid-fire manner that is clearly raising questions. Note: this is nothing against the linking philosophy that Tony et al have about scarcity of links as I do agree with that approach; this is about using tools in rapid succession to impose that when I know it doesn't yet enjoy far-and-wide consensus. This is exactly where BetaCommand got ridiculed by the community, and if such actions continue, it will happen to those rapid-fire removing links for the same exact reasons. --MASEM (t) 12:57, 26 April 2012 (UTC)
<sigh> What do you mean by this emotive "rapid fire"? I certainly don't engage in it. Bear in mind that gnoming and its timing is more complicated that you're making out. For example, a few thousand article-stubs on Polish villages alone have been dumped on en.WP from the Polish Wikipedia (foreign dumping is becoming an increasingly common phenomenon—it is not an exaggeration to call it a waterfall, and other WPs have different standards of typography, date-rendering, and wikilinking, not to mention other issues); each Polish village article uses almost the same formula, including links to "village", "region", and "population"—well, that's what they do in most WPs, senselessly. Gnoming is multilayered, at least as I do it. To start with, working out which date format to harmonise to is a pain that requires the whole diff to be looked through. Then you have to check for the holes that might have been created when geographical names are listed inline (an infrequent issue, but nevertheless it has to be checked). Some articles are stubs; others are more substantial. Occasionally one fixes glaring prose issues.

This mantra of yours, "there is no deadline", has no place in Wikipedia: it is a cop-out, a call to go slow, the very opposite of what is needed at the moment. It ends up being very damaging, particularly when it's used to discourage hard-working editors who appreciate the massive task of article maintenance—a task we can never keep up with, but try to. And the mentality it pushes is getting the project into trouble when it comes to managing some of the serious problems we now face (buzz me and I'll enumerate them). The neglect of PR and company people who do the right thing by asking for changes to be made to an article (thus avoiding CoI edits) is just one current issue where "there is no deadline" comes up against the rigours of real life out there, where a powerful wiki can cause a lot of trouble when bias and factual errors go uncorrected. So please, re-examine the damaging potential of your mantra and avoid splashing it around: it harms a project that needs volunteers to embrace professional standards to survive. It's no longer 2004, when such statements might have been bandied about with less egregious effect. Tony (talk) 13:45, 26 April 2012 (UTC)

I strongly recommend you read the whole of Wikipedia:Administrators' noticeboard/Δ, including all archives, and Wikipedia:Arbitration/Requests/Case/Betacommand 3 including all case pages. My concern is not the philosophy of what should be linked, but the manner to enforce it. We are not where Beta was yet, but the conditions are ripe.
As to the deadline point, there's a huge difference between COI and patently false information being posted on articles about companies and people, and dealing with a sea of blue links. The former could get WP in trouble and needs to be dealt with post haste, but the latter is an annoyance but far from any legal ramifications for the foundation. The Foundation has made it clear there are three areas where there actually is a deadline: Bios of living people, non-free media, and copyright violations - all those feed back as potential harm to them. Anything else is style and content, and can be fixed as volunteers get to it. Hence, as long as we remain a volunteer project, there is no deadline to correct annoyances like over/mis-linking. And it is better to do it in a manner that has consensus than to heavy-hand it as proven by Beta above. --MASEM (t) 14:10, 26 April 2012 (UTC)

You've side-stepped my point, again, by sliding in a binary attitude now: oh, just three things are urgent, so the rest can be cast in your call to go slow, your "there is no deadline". It would be nice if the world were cast in simple binaries, but it's inevitably more nuanced and multilayered than that. So I take it you intend to persist with your "there is no deadline" ethic; it's very damaging, and I wish you'd rethink it. Tony (talk) 15:02, 26 April 2012 (UTC)

Sometimes Masem writes as if only "rapid fire" AWB use (like Betacommand) is his concern. Sometimes, as in "Until we have a casebook" (i.e., when pigs fly), he says there should be no AWB country delinking at all unless one WP:OWNs the article. It makes a difference. Art LaPella (talk) 14:22, 26 April 2012 (UTC)
The argument that there can never be a casebook is bogus; the problem is that those employing tools like AWB are following a casebook of their own desire and not one that has been determined by consensus (except among a select group). And I will stress again: I probably agree with most of the linking advice that is in those personal casebooks. I just find that the community does not accept the use of semi-automated tools to enforce something that does not have wide consensus, as is what is happening here.
There is a way to write the casebook, just like for NFC we've had to write a non-inclusive lists of acceptable and unacceptable uses. Take country names: I would agree that if we actually had a RFC to say "country names should not be linked except on articles dealing with geography or political issues" to make that a case, consensus would back in, and from there, AWB delinking of country names from every other type of article is fine. The problem is that doing the AWB steps before getting consensus on such cases is putting the cart before the horse, the same problem that Beta repeated came under. The community does not accept that type of logic - that's why bot operators have to get approval via consensus and usually undergo strict penalties should they do something intentionally outside their scope. Now, with automated tools like AWB , its a bit different, but the same logic and courtesy should be made to the community to make sure that the task that is being rapid-fired is what the community accepts. Delinking as been evidenced in this discussion so far, as best as I can tell, does not yet enjoy proven consensus. Set that up first via an RFC, and then you can delink as fast as AWB allows you to for those specific instances. --MASEM (t) 19:40, 26 April 2012 (UTC)

Even after an RFC, I would not countenance "rapid-fire" AWB; the decision-making is too contextually based, which is one reason A di M has said "pigs might fly" WRT your case-book proposal ... it's not easy to create a catch-all list, since there's always going to be a grey area, even though much of the matter is clear-cut. Who's using AWB???? How fast does AWB go? Tony (talk) 07:05, 27 April 2012 (UTC)

Art is the one that brought up AWB, but it can be any user-created script that is of concern if mis-used. AWB and these work as fast as generally the person running the script can hit accept and have en.wiki acknowledge the change. But, as per the Beta case, the problem is that when there is very little time between these script-enabled changes, the amount of human checking involved with these changes are raised into question. As an example, you today, around 06:30 on April 27, had about 8 edits over a small time window (<10 min) using some automated tool (based on the change message). Now, that rate of editing is completely possible by a human if they are making simple changes, but with the bulk of changes being put into place via these diffs, its hard to say for sure if you're double checking that each change is correct or if you're assuming your script is correct and just clicking off on the change. That itself is not bad - for comparison, Beta was one limited to an average of 40 edits in a ten minute block, which was considered the bare minimum for a human to check the results of an automatic script and accept it to en.wiki. You're nowhere close to that, obviously, but the same idea does apply: when you rapid-fire edit with an semi-automated tool of either your own design or like AWB, the human check of those results are called into question. The example given elsewhere in this thread where geographic terms were delinked from an article about geography is the example of human failure to catch this, and it was part of a rapid-fire block.
As to the casebook issue, you're just providing the counterargument for why links shouldn't be taken out by these semi-automated tools because their value is contextually-based. An semi-automated tool cannot figure that out. Now, again, if it is the case that a human reviews the suggestions of the tool, that's fine, but then the rapid-fire nature and whether there is truly human review of each link removal comes into play. If there was at least a case book that affirmed through consensus the types of articles where country names should be and shouldn't be linked, then this issue of rapid-fire changes becomes less an problem, since now that's acting upon a consensus-affirmed specific type of removal. Right now, the one line about what generally shouldn't be linked doesn't cut it. --MASEM (t) 13:46, 27 April 2012 (UTC)

Please be more careful before you plaster accusations around. You are accusing me of carelessness. Tony (talk) 08:03, 28 April 2012 (UTC)

There is community consensus for using AWB without needing approval for every edit. There is consensus (nobody has ever specifically asked me to get BAG approval), but paradoxically less consensus, for using AWB in compliance with the Manual of Style, even though the Manual is evidence of consensus. It's like a policeman who keeps threatening to arrest me for not breaking enough laws. I haven't used AWB much since last year, and maybe I shouldn't until everybody comes to a coherent understanding on this issue. Art LaPella (talk) 01:45, 27 April 2012 (UTC)
But it is expected that when you are using AWB for a large number of edits, those edits are the type that have wide consensus. The problem is that the type of links being removed are not spelled out as improper in the MOS. Do an RFC, set it as a casebook example, and then its not a problem. Until then, such edits via AWB or whatever scripted tool will be considered as against consensus and a problem, and as per Beta, dealt with appropriately. --MASEM (t) 04:33, 27 April 2012 (UTC)
"the type of links being removed are not spelled out as improper" implies that it is possible to use AWB to remove country links with your approval. But often you seem to say the opposite. Easily fixed: do you condemn all AWB country delinking, or some AWB country delinking? Art LaPella (talk) 05:34, 27 April 2012 (UTC)
Since there's nothing about country links in Wikipedia:Manual_of_Style/Linking#Specific_cases, they should not be removed by rapid-fire AWB edits or similar scripts since there's no established consensus to remove. This can be fixed by having an RFC to get consensus that country links shouldn't be linked outside of geography/political pages. --MASEM (t) 05:42, 27 April 2012 (UTC)
I wouldn't call my AWB edits "rapid-fire" (it handles a long list of Manual of Style issues, and I'm likely to pore over each article up to an hour), so let's ask it this way: do you condemn my AWB country delinking? Art LaPella (talk) 14:13, 27 April 2012 (UTC)
It is likely not a problem since yes, you appear to be spending time to assess the links. (Again, I agree generally countries shouldn't be linked in most articles). Its when others spend under a minute on a page to remove such links that brings into question if there's human review of the changes. Again, I urge reading of the Beta/Delta cases to understand the community's concern with rapid fire semi-automated editing. --MASEM (t) 15:35, 27 April 2012 (UTC)
Thank you; I'll save a link to that comment. Although admittedly, if I were forced to spend more than a minute contemplating each country link, I would probably have to spend most of that time looking out the window. But at least the demand for more time between edits is a frequently heard one. Art LaPella (talk) 16:57, 27 April 2012 (UTC)
The case would be different if there was clearer advice in MOS to say when and where country links should be avoided - which I totally believe we can spell out with allowable exceptions - that would make the more-automatic removal-like (read: rapid-fire) of country links be acceptable to consensus. Not to belabor the point too much, but the background on Beta's case is critical to understanding what I'm trying to get at here. --MASEM (t) 17:16, 27 April 2012 (UTC)
  • No, I've just flicked through the Betacommand thing: 80–125 edits in 10-minute periods, apparently breaking pre-existing sanctions, civility and engagement issues: hello? What are you banging on about? Now you need to stop castigating and bullying editors for doing careful work in cleaning up articles according to the guideline. Tony (talk) 02:36, 28 April 2012 (UTC)
    • If you don't understand how this situation (extending from date delinking to the article titles ARBCOM cases) can potentially grow into the same as the Beta case, you probably need to spend a bit more time understanding the issues. Yes, in Beta's case, there is a contributing factor of Beta being a problem himself in attitude. But take that out, and you're still left with the idea that rapid-fire edits where no clear consensus exists will be seen by the community as a problem. Either you edit like Art where there's clearly time enough to show careful evaluation of links, or you get a community RFC to establish what cases of delinking you can do by rapid-fire semi-automated tools. The reason, particularly with the MOS and following from the ARBCOM cases, is that a small number of editors here will change the MOS without seeking wide consensus - even if these are completely sane, reasonable choices - and then start editing rapid-fire as if this was always the away. That's a problem if the community doesn't agree with that change. That's what started this whole thread. Again, I'm not telling you you can't delink countries. Just don't do it in a rapid-fire manner unless you know the community consensus is behind specific removals. --MASEM (t) 03:27, 28 April 2012 (UTC)
  • This latest pronouncement suggests that you're clinging to a deep fear of anything that moves at more than a snail's pace on en.WP: this is the only possible interpretation, I think, of the dictum you're clinging to, "there's no deadline". Now, there is consensus, as someone pointed out above, for avoiding the linking of major geographic locations—it's in the style guide, and the style guide is the outcome of a somewhat stormy rationalisation and set of compromises brokered by Kotniski in late 2009, I think it was. Like dates and months and centuries and days of the week, I've seen virtually no issue taken by the community over not linking "Australia", "United States", "Germany", "China", "Russia", "New York City", "Los Angeles", an so on. We do assume that our readers have a mental age of more than five. And let me say that a large proportion of country-name links are lazily inserted when a more specific link would be so much more helpful to readers. We cry wolf by being cavalier in linking; caution will earn the linking system a greater functionality. And also, WP is not a dictionary, as the first pillar says, so if you want to look those items up because you don't understand them in the context, Wiktionary is the place to go. Thanks. Tony (talk) 06:24, 27 April 2012 (UTC)
    • I am not disputing, and in fact agree, that links to major geographic names are useless on most pages. That's not in dispute. The dispute is when tools like AWB are being used in a rapid-fire pace to remove such links on pages where they are appropriate, such as topics about geographic features (and examples where that has occurred been identified in this discussion). That is careless, and that's why we should not be rushing this. Again, stressing this: it is not the issue about what actually are the best types of links to be made and when links are poorly made - it is about the fact that editors using rapid-fire tools seem to be following personal, non-consensus decisions on what should be linked to mass delink terms. If the advice to remove certain links are not present on the MOS page, that should not be done by mass edit tools in a careless fashion. Establish the concept of scarce links that are the most germane to the article (a concept I completely agree with) with specific examples and cases via an RFC, and then mass-edits to enforces those cases won't come under any scrutiny and can proceed as fast as you want. Right now, without hard specific advice, rapid-fire edits will be seen as harmful by the community, as evidenced by Beta's case. --MASEM (t) 07:00, 27 April 2012 (UTC)
    • Traditionally, geographical place have belonged more in encyclopaedias than in dictionaries, so that particular argument isn't terribly strong. If you weren't familiar with Tashkent, would you rather see this or this? ― A. di M.​  11:18, 28 April 2012 (UTC)
      • Tashkent is a strong candidate for linking in many instances—although perhaps not just after its capital city has been linked, unless it's a rare case in which there's a reason to bunch-link a specific and a general; and perhaps not if "History of Tashkent" would be more specific, in which case you'd try to expose the angle of that target in the grammar of the pipe, not always possible, I admit. "Australia", "UK", "US", "Canada", "Italy" and the like have far weaker claims to be linked; but the reason one scrutinises diffs if using a script is because there are very occasionally reasons to do so. It is, after all, the English WP, not the Tashkent WP. Tony (talk) 11:39, 28 April 2012 (UTC)
        I think you didn't get my point – try reading it again. (I didn't mean that Australia should usually be linked; I meant that the reason why it usually shouldn't has hardly anything whatsoever to do with WP:NOTDICT.) ― A. di M.​  20:48, 28 April 2012 (UTC)
Tony: we are a volunteer project. Since there are no time demand requirements by editors we cannot expect any time requirements for articles to become ship-shape. Ergo, "there is no deadline" has to be how en.wiki operates, except where the Foundation has identified potential legal problems for themselves. Alternatively, let's put it this way: okay, removing excess links is a necessary help for articles to become easy to read and comprehend without being lost in a sea of blue, which is a point I would agree with. But then why stop there? Why not copyedit the whole article, since that also helps? Why not fix all the references, find all the missing ones, fix the punctuation, order sections appropriate, assure that dates and US/UK spellings are consistent, etc. etc. all things that also help comprehension of an article? All the semi-automated scripts that I see removing links are doing some other mechanical things (like stripping unused params from templates - a good thing) but if we have a deadline, why stop there? We should be copyediting and improving articles every time we touch it with AWB scripting too, since obviously comprehension must be established. Of course, I hope you recognize that this is a bogus argument, but that is what you are basically saying when you say "WP:DEADLINE is harmful". It's not - it's how the wiki works - articles will get improved in time as fast as volunteers can do it. An article that is of poor quality - whether due to excess links, bad english, inconsistent formatting, or whatever - does appear to be a problem, but the Foundation is not going to be legally liable for any of those faults, and ergo, it is not a high priority to fix them. That's why DEADLINE is critical to keep in mind, and why rushing off to try to delink terms without seeking wide consensus is the problem here. --MASEM (t) 19:40, 26 April 2012 (UTC)
I only noticed this section since I posted below, but I admit that I stopped reading "Linking sense" after this point: "Granted gross overlinking is stupid and sets teeth on edge, but it is rare and otherwise essentially harmless, venial". It is not "essentially harmless" because gross over-linking dilutes the more valuable links that assist the reader in gaining a deeper understanding of the article. The fact that "it is rare" is because of the hard work that dedicated and experienced editors have put in over the past many years to reduce the scatter-gun approach to linking that used to be endemic at WP. I don't run scripts, but I do a lot of article work in which I reduce over-linking, and with the exception below (which is a delinking that I didn't instigate, but support), very rarely receive any resistance to delinking from local article editors.
If you want a taste of what happens when "we drop most of the strictures on formally functional linking" and "leave it to good sense...", have a look at the scatter-gun approach to linking that is now the norm on the French Wikipedia. Here's the link to today's FA there, which contains eight links in the first sentence—none of which help to deepen the understanding of the article's subject. Note how, besides his name and the word "américain", every word over three letters long has been linked? Also note the bizarre attempt to link "1958" to their "1958 in cinema" page—as if what happened in cinema in the year the article's subject was born is in any way going to help readers gain a deeper understanding of the subject. And that's one of their FA?
From the coalface ... please don't weaken the current policies because (in my experience) they have met with extremely wide acceptance, and have definitely helped to give our articles a more professional and reader-friendly appearance.
GFHandel ♬  01:26, 26 April 2012 (UTC)
Tony, I see you are still making comments like this after all this time, alleging that I and others want "maximal linking". Unless you strike or correct that, it is clear that you are either extremely thick, cannot read or are a liar. The fact that your post opens with an allegation directed at another editor that they are "twisting things" truly takes the biscuit, as the cliche has it. And the gall of someone who complains about a "merry-go-round" and wanting "something new", when it is they who for years now have charged through pages with their battering-ram script and refused to address points made to them is something else. You know, if you actually listened to people and made a substantive acknowledgement of their points instead of making up crap in response, you might not find them having to repeat themselves. You can blame it on the usual suspects in your head all you like if it makes you feel better, but doesn't the succession of new names coming here to query what you do raise any doubts in your mind? The regularity of names is far more obvious in the small number coming to defend this blanket delinking. And yes, I am still waiting for a) evidence of consensus mandating such wide script-based auto de-linking; and b) evidence for the claim, now also made by old friend GFHandel, for "extremely wide acceptance" of such actions post-facto. I could at least stop asking those questions, if you'd ever deign to answer them with anything other than assertion.
To clarify, for you (for the 406th time) and anyone else, I broadly support much tighter and more focused linking - including the removal of links that repeat in close succession, links to common English words and links to proper terms when those things are not directly relevant to the main topic at hand - and doing that as part of the normal copyediting process. There may or may not be consensus for that level of restriction of links; people will certainly have different opinions on it, different from mine as well as yours (as they would seem to in France, when it comes to FAs). However, I do not support blanket removal of links to specific terms from thousands of articles, regardless of context, on the basis that they are supposedly "well known" things by some unidentified standard. Nor is there explicit consensus for you, or anyone else to do that. Use a script, sure, but go back and reinstate relevant ones (you say you do this, but I rarely see it). I've lost count of the number of times I've seen scripts strip out perfectly useful and relevant links to supposedly "well known" terms from a page, while missing out and leaving far more useless, repetitive and redundant links sitting there. Manual and discretionary editing would have made a much better job of improving and focusing links, while at the same time offering decent and broad navigability for readers to related and often quite significant other articles in an online encyclopedia. N-HH talk/edits 07:15, 26 April 2012 (UTC)
@GFHandel, re the France example. FWIW, personally I am fairly open-minded about linking or not when it comes to a string of nationality/professional terms, eg "X is an American actor, singer, model ...", and am sympathetic to objections to it (the infobox to me has always seemed a better, alternative, place for such links). But I think that the fact that French WP appears happy with it, while not directly relevant to establishing consensus here, does of course indicate that there are a range of views - even if they are different to your own - as does the fact that many editors do commonly link in that fashion, even if they don't speak here (and btw, the fact that people rarely complain about delinking isn't evidence of consensus or approval - it's more likely to be evidence of "meh" or an erroneous assumption, especially with script delinking, of some quasi-official drive). Also, I'd take issue with your focus on links that "help the reader understand the topic". Sure that's one function/purpose of linking, but not the only one. As the guidelines say at the very start, links "bind the project together" - ie they provide navigability. The problem is that not only have the guidelines been chipped away over the years by people with pretty determined views (rather than being the definitive expression of specific consensus), but they still don't even say what many of those people want or believe them to say - navigation and connection remain criteria for linking; as does relevance, even for common terms. N-HH talk/edits 08:43, 26 April 2012 (UTC)

GFH, The coalface is a broad one, accommodating many workers and many kinds of work. Let no one claim it for his personal fief. I am such a worker on several kinds of work, and I do not doubt that you are one, nor that all the others contributing to this section are too. Whether the "current policies" have met with acceptance is not what defines their validity in all contexts, otherwise they would not be "current", but cast in concrete. "Weakening" the policy is not an issue, and not a consequence of rational discussion. I am nonplussed at your suggestion that it might be. Perhaps you would like to enlighten any readers who fail to follow you? I am objecting to difficulties arising from the attitudes of certain classes of editorial behaviour, and that is not a matter to be settled by tarring every proposal with the same brush just because we can instance a few horrible examples. For material reasons I insist that over-linking is worse than under-linking and ultimately "essentially harmless", and in fact the reasons include those that I instanced in the rest of the list at the start of this section. If you reject material reasons unread and out of context because you have in anticipation favoured what you fancy to be an opposing reason, that is up to you, but it hardly contributes to establishing a sound perspective on the matter. Over-linking is essentially harmless because it is a passing irritation, and one that can readily be fixed by the irritated person if s/he is an editor. If s/he is not, then it is unlikely to be a bother after the first few minutes, if that, because anyone intelligent enough to read and to consult WP should be able to work out pretty quickly what the blue text means and, after a few trials, what the red text means as well. Conversely, under-linked material can mean that one does not realise that there is a link to search for at all, or if one feels like going exploring, one does not realise what title it is available under. You might find it helpful to reflect that the same blue-tinted word in the same article can validly and helpfully link in different places to different explanatorily linked material. Precisely what you imagine "formally functional linking and good sense" might mean, puzzle me. You appear to imply that it doesn't matter how one links as long as it is only according to the principle of a single link at the first appearance of a highly obscure term in an article, and that the only alternative is a French mess. I haven't checked on your horrible example, not because I doubt your word that it is a horrible example, but because even if the editor had idiotically linked every single word in the article, it would not have been in the slightest relevant either to good sense or formally functional linking, because it would represent neither, and to the reader might well be irritating in passing. Nor would leaving out valuable links be any better, and it would let down the reader in need by causing greater disruption if he went searching for links, and very likely depriving him of necessary information if he did not, or searched unsuccessfully. Furthermore, the reason for a link in one place might differ from the reason for the same or a similar link in another place, in which case it might well be important to link both. Linking might or might not be red, in which case its presence would be important both to the reader and to the appropriate editor, as opposed to blue, in which case it would not generally attract attention from the editor, but would leave the reader free to choose. If you fail to agree that these are matters of good sense and formally functional linking design, then please explain as cogently as you can manage, just why they militate against the sensitive and practical placing of links where they are constructive and omitting them where they are not, and I promise that I shall read your views as carefully as you read mine. Or even more. JonRichfield (talk) 09:44, 26 April 2012 (UTC)

Anyone care to look at Gisele Bündchen?

Latest comment: 12 years ago2 comments2 people in discussion

Hi. Could someone experienced in matters about linking please see if I am on the right track here? I think it would also be nice if a MOS-compliance script could be run on the article. Thanks in advance for any assistance possible. GFHandel ♬  00:18, 26 April 2012 (UTC)

I think it should link to German Brazilians which in turn should link to Germans and to Brazil, but not directly to the latter two. ― A. di M.​  10:36, 26 April 2012 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Manual_of_Style/Linking/Archive_13&oldid=1258126239"