Wikipedia talk:Manual of Style/Linking/Archive 15

Archive 10

←

Archive 13

Archive 14

Archive 15

Archive 16

Archive 17

→

Archive 20

Confusion over "particular relevance" and major cities.

User:Tony1 has recently trimmed several Wikilinks from Bill Nye, citing overlinking. One term was Nye's hometown, Washington, D.C. Here, it says Unless they are particularly relevant to the topic of the article, avoid linking the names of major geographic features and location. But then Tony1 removes it with this contradictory statement, and now another editor is reverting my reversions, claiming only that Tony1 has explained it. Can another editor here clarify? If a hometown, where a fourth generation resident lived till college, is not particularly relevant, what is? In what case could Washington, D.C. be "properly" Wikilinked? And if major cities are not subject to the relevance exemption, why does this guideline specifically say they are? InedibleHulk (talk) 01:14, 31 October 2012 (UTC)

Don't distort the facts: you're up against several editors who've come out of the woodwork to revert you. You've offered no cogent argument on the talk page. And a "proper" context might be a geographical anchor article, among others; certainly not just because some guy was born there. Oh, and Bill Nye could be licking the soil in Washington DC: it would make no difference. Fourth-generation? Why not link to the 19th-century history of the city, then, if you're worried about that? Tony (talk) 02:43, 31 October 2012 (UTC)

My argument has been pretty clear in the edit summaries. It's relevant to Nye, and not in violation of WP:OVERLINK. But this isn't the place to argue. I'm looking for clarification on the rule from another editor, since we can't seem to agree on interpretation. InedibleHulk (talk) 03:28, 31 October 2012 (UTC)

It makes no difference anyway; Washington DC may be well known in the United States of America, but I bet you that half your non-American readers (and a surprising number of American readers too) couldn't tell one Washington from another; quite a few couldn't even tell you why they are called Washington, and if they could, couldn't tell you why Washington isn't the capital of Washington, or even why there is a DC when it isn't capitals of electricity and what it means (or for that matter, why New York isn't capital of New York). How many of such, who regard themselves as Mpumalanga could even tell me offhand what it still is the the capital of, or how long it has been so? I do realise the deep pride, and far be it from me to deride it, that many Americans take in their own provincialism, but to expect roughly 95% of Wikipedia's potential worldwide readership to be fully informed in trivial matters of tricky local geography, has all the arrogance and none of the dignity or even the excuse of the Duke of Wellington's "Let 'em learn English." In context Washington, D.C. most decidedly should be linked, just like Greenhorn. Over linking doesn't even come into the discussion. And yet there are some here present who sniff at terms like "vicious unlinkers", and have the nerve to prate about the art and science of intelligent linking! JonRichfield (talk) 14:49, 31 October 2012 (UTC)

My thoughts exactly (in different, fewer words). InedibleHulk (talk) 16:27, 31 October 2012 (UTC)

We expect and design en.wiki to assume a high school level of English proficiency as a lowest common denominator. Regardless where you are in the world, if you have that, it's extremely hard to imagine the situation where you don't know what Washington DC is, much less where cities like Cairo, Frankfurt, Moscow, Cape Town, San Paulo, Shanghai, and Tokyo, among others - all major metropolitan centers of major countries. Linking to these cities just because they are a proper name but with otherwise little relevance beyond a point on a map dilutes the "What Links Here" for these cities, and weakens the encyclopedia. This is why links to major geographic features in articles unrelated to concepts like geography or politics/economics are strong discouraged. --MASEM (t) 15:38, 31 October 2012 (UTC)

Sure, most English speakers know of Washington. It's where the American government works, and contains the White House and the Washington Monument. But there's a lot more to it than that, which most outsiders probably know nothing about. If someone's interested in Bill Nye, they may want an idea of the place he grew up in for twenty years (its education and media systems, for instance). In this article, it's far more than "just a proper name". Seattle is linked immediately below Washington for that kind of reason, and nobody's complaining about that. InedibleHulk (talk) 16:21, 31 October 2012 (UTC)

Which is why we should be linking the city in the infobox (where the OVERLINK on geographical terms is not applied since it's not prose). I'm surprised its not (and doesn't appear to be part of Tony's change). But linking something that's common in prose is extremely distracting and not helpful. I probably would not link Seattle directly, either, since it is a well-known major metropolitian center, but that's less a clear answer. (I do know that with the appropriate links on Nye's page, you only have to click once to get to a page that has a link to Seattle). As for things like the influence of schools and media on a person, that's why we can link in high schools and shows/concepts the person's worked on, which are a lot more germain to the person than the city those are located in. --MASEM (t) 16:35, 31 October 2012 (UTC)

Stop press! I agree! (Sorry 'bout that!) I agree about linking to things in infoboxes etc. OK, re-start them presses and we can get back to boring disagreement. Links in prose are by no means distracting after the first ten minutes in the medium, and in particular they are not nearly as disruptive as encountering things that one doesn't happen to know, or if one does, doesn't know what aspect or attribute the author had had in mind, or any of the other reasons I mentioned for linking. So sorry if that misleads you into thinking that I might contradict you, but I just had to correct your flat statement of a personal, matter of opinion that you represented as a matter of fact, when it is a matter that either way affects hundreds of thousands or even millions of readers on your and your allies' say-so or say-not. Oh Reason! Thou art fled to ... Where your view differs from my deeply-thought-out font of truth, is that mine offers the readers more choice, particularly if the improvements that I proposed were implemented. Which incidentally were never substantially rebutted except by equally flat statement. Coincidence? JonRichfield (talk) 18:06, 31 October 2012 (UTC)

It (and Los Angeles) were delinked in the infobox by a bot run amok a week ago, shortly before Tony showed up. I restored them, and they were re-removed. Just curious, why are major cities more appropriate in geographic or political articles? Wouldn't someone reading about geography or politics likely be more familiar with places like this than the general reader? InedibleHulk (talk) 16:44, 31 October 2012 (UTC)

When talking about smaller towns or other nearby features, a link to a nearby geographical feature may provide additional details that may not be there in the smaller town article. For example, Oak Park, Illinois would benefit from links to Chicago where more detail about climate, population distributions, politics, media, arts, etc. would all be present but that would not be covered on that smaller community page outside of the local aspects. --MASEM (t) 17:05, 31 October 2012 (UTC)

<Siiigh!> Masem, your imagination and its failures are of course definitive criteria. Pardon my having overlooked them. I didn't see them linked anywhere. I grant that not everyone in the US shares the world-wide knowledge of the (very friendly and genuinely nice to meet and often obviously intelligent) US citizens who, having asked where we came from, and told the country name, responded: "Oh, what country?" Note that this is not a sneer at them; I couldn't list more than say half the world's countries offhand, and heaven knows how long it would take me if I tried. As for criticising anyone who couldn't call my home town to mind... spare me! It all is however an indication of how many of the world's prominent place names (not even counting some of the smaller countries and cities) simply get swallowed up in the basis for the Chinese indifference theorem. And of course, I totally had overlooked the peril of diluting the power of the "What Links Here" for major cities. How silly of me! Have you lately checked how many link to "Washington" or "Washington D.C." or "Washington (state)"? I tremble to think of the accumulative consequences of yet another link to any of them. It would probably bring the whole system down, wouldn't it? Let's calculate the marginal effect on the "what links", of each new link on each of those three for a start shall we? And then let's compare the cost in comparison of the benefit for each user who clicks on the link, hm? But never mind that shall we? How about rather let's nominate WDC as the smallest city that definitely requires no linking. It then would argue a rather chauvenistic attitude to permit linking any larger city then, wouldn't it? How many of the sixty-odd larger cities would you undertake that half the English-speaking population of the world (not just the US!) would recognise offhand, let alone be able to place regionally or in general significance? I don't know you personally, so you might be a memory prodigy or a professional geographer or demographer, but what odds would you think it would be fair for me to offer that you would be able to place two thirds of them without doing some looking up? For example, by clicking on a link? And for Joe Average High School English Reader? Just wondering... JonRichfield (talk) 17:44, 31 October 2012 (UTC)

It's not WP's fault if you aren't taught basic world geography in an English-based grade school as a requisite for high school - or if you forgot. Yes, there probably are some countries that are not common that I couldn't easily point to on an unlabeled map - several African, Middle Eastern, and Southeast Asia ones, for example - but at the same time there are terms that are stables of world geography that not knowing is likely ignorance more than schooling. Remember that we're aiming for the high school level of English speaking populations, not the entire world, where education systems are generally some of the best in the world. Even where English is being taught as a second language, I'd still expect that they would be given instruction on the names of major countries and cities around the globe in English.

And yes, dilution of "What links here" is very important. You can't say "oh, just one more link in that list will hurt", because the problem is right now those links are far too diluted. There's nothing to ruin because it already it. But if we had much better selectivity of linking to cities, the What Links Here would be very important and useful as a research tool, and links added without thought would dilute that. --MASEM (t) 18:01, 31 October 2012 (UTC)

"I weep for you the walrus said, I deeply sympathise…" Masem, you really shouldn't reply so promptly and unthinkingly; the results are painful to sensitive spirits. "It's not WP's fault if you aren't taught basic world geography in an English-based grade school as a requisite for high school - or if you forgot." That is your version of vaunting of your education system is it? Recalls: "I am the product of this college; what I don't know isn't knowledge" hm? Professor Jowett would be deeply affected. I did after all speak of the English speaking world, didn't I? Did you imagine that it ended at sea and shining sea?

And of course the 49th parallel...?

"Yes, there probably are some countries that are not common that I couldn't easily point to on an unlabeled map - several ones, for example" …

Oh well, don't fash yourself please; who cares about uncommon little backward countries like African, Middle Eastern, and Southeast Asian, let alone their cities, hm? With their cities all smaller than Washington DC? Not counting some sixty or so? After all, they have inferior, non-English education systems don't they? Not at all like yours, hm? But what has that to do with your educational accomplishments related to whether one links to those countries or not? I am no geographer myself, but if I were interested, I could click on a link to each and find out all sorts of fascinating things about either or both, (did someone say serendipity?) if rabid delinkers did not get to the articles first. But it would be a lot easier to find out what was worth finding out if they were in fact linked, don't you think?

"... at the same time there are terms that are stables of world geography that not knowing is likely ignorance more than schooling. Remember that we're aiming for the high school level of English speaking populations, not the entire world, where education systems are generally some of the best in the world. "

Oh I see. The reason that you now are showing up so well that you don't need links to items that lesser mortals don't know offhand, is that you have benefited from your uniquely superior English educational system? Or is that it in spite of the lucidity that your English education must have instilled, you have misstated your intended meaning, and really think that the "entire world, where education systems are generally some of the best in the world" is so much better than the English system? I assume that for example you casually learned all about all sixty or so <ahem!> stable cities larger than WDC in your superb schooling and why they are larger, though less important and less interesting? And why should WP cater for anyone lesser? I don't suppose you ever bothered to ask yourself what the reasons for the putative superiority of your education system might be? Do tell us when you have worked that one out!

Even where English is being taught as a second language, I'd still expect that they would be given instruction on the names of major countries and cities around the globe in English.

Now, now! Please don't burden the lesser breeds without the law with the educational advantages that you so nobly bore up under!

Certainly as novel an outlook as trenchant though. Well done!

You helpfully explained: "And yes, dilution of "What links here" is very important. You can't say "oh, just one more link in that list will hurt", because the problem is right now those links are far too diluted. There's nothing to ruin because it already it. But if we had much better selectivity of linking to cities, the What Links Here would be very important and useful as a research tool, and links added without thought would dilute that."

Garbage Masem, garbage! What you seem unable to recognise is that the function that you wish to allocate unilaterally to linking is not amenable to such a process, not with any precision anyway. It wasn't designed for it and if it never got used for it but did get used for linking (remember linking???) that would be fine. Remember that your argument works both ways; a missing link would be just as harmful as a redundant link, and a redundant link for one purpose would be a vital link for another. And the more links the less important any new additions or deletions. If you were any good at systems design you would recognise such facts on day 1 and would design a suitable tool and stop trying to wreck a far more important system than your own research whims on a hiding to futility. Or you would design a tool to piggyback on the working linkage system and do some useful winnowing. Guess what my bet would be on the probability of your doing either?

Do let me know when you get that one worked out too. JonRichfield (talk) 19:58, 31 October 2012 (UTC)

I'm being serious. I cannot imagine, even in the poorest of countries where English is taught as the primary or secondary language, that students are not given a basic understanding of world geography. They're certainly taught other common English words that would would never consider linking outside of extremely germain uses (eg cat, dog, house, with "cat" being used on pages about the feline species). Well-known proper names of countries should be treated in exactly the same fashion, and I find it very hard to have someone well trained in English to not recognize what Washington DC is.

We are a research tool; we need to cater to the average reading style and expected education level. A low density of links is favored in this form, otherwise they are distracting - both in being blue text as well as tempting targets to click on. We should be striving only to provide the links with the highest research value to the topic at hand such that readers know that when they click a link they are being taken to a highly relevant topic to the one they are presently reading. When we introduce (in prose) low value links - like the link to Washington DC on Bill Nye's article, along with high value, readers will start to become less inclined to follow links if the quality of the links are so varied.

Yes, there are going to be English speaking, high-school trained people that don't know what Washington DC is, but that number is a small minority. We can't cater to these people without ruining the work for the rest of the readership, which is why we have chosen the above lowest common denominator for our prose. --MASEM (t) 21:21, 31 October 2012 (UTC)

I too am being serious. That is where we differ. Who said that a "basic understanding of world geography" is only basic or only an understanding, if it includes all the distinctions concerning places and people called Washington, in such detail that they neither need nor want a link, irrespective of the context, even if that is mainly in your superior Anglophone educations? Why pick on Washington? There are over sixty larger cities, including older, more stable, better-known, and important ones like Paris, London, Moscow, Los Angeles, New York, and Mecca. And that is totally without even addressing the question of when to lower oneself to linking to an important article. Why you should want Washington D.C. (for example) to be an orphan article just for the sake of your research, escapes me. You must give us a run-down on that some day. Don't bother to address to TDLR-Tony though; waste of bandwidth. He is a small minority; don't worry about hurting his feelings. Don't even consider including links for minorities in general. Let 'e learn English -- your English. But Do explain what this vital research is that you are undertaking, research that justifies highjacking and sabotaging a really important tool, and explain why it is so important for that research to have a really large number of orphan articles to work on. JonRichfield (talk) 13:30, 1 November 2012 (UTC)

Given that this is "en.wiki", and that there are many other language-specific (note: not geographic-specific!) wikipedias, the presumption that we expect a reasonable understanding of the English language is not elitism; otherwise, we would have to cater not only to those with a loose grasp of English, but also to the younger crowd that are just starting to learn English - what the Simple English Wikipedia attempts to do. That would make it incredibly difficult for us to do any type of reasonable coverage of a topic. So no, the "elite" expectation of a good English background gives us the right legs to actually build useful articles.

The reason we highlight major geographical features like countries and major cities is that in casual English conversation, these more often than not are considered common words, akin to words like "cat" or "house" outside of being proper nouns; our expectation of an English background would imply that the reader should be aware of these as nouns even if they can't pinpoint where it is on a map, just like we wouldn't expect the reader to know the biological details of a cat just because it's mentioned. Thus, linking these should be limited to when the use is highly germane. Take "cat" for instance. In a bio article, linking to "cat" in "John Smith keeps a pet cat." would be a wasted link - we're not asking the reader to understand what a cat is in any detail. On the other hand, statements like "Egyptians were one of the first to domestic cats" or "A panther is the member of the cat family", now the term "cat" has strong direct relevance to the discussion at hand, and a link would be reasonable for the reader (though I think in both cases, there might be more specific links that would be better, but you get the idea). Similarly, for Wash DC, "Bill Nye was born in Washington D.C." doesnt require any specialized knowledge of DC to understand, and thus a link would be a waste, but statements like "The Potomac runs through Washington D.C. before emptying into the Atlantic" or "John Smith was the mayor of Washington D.C. from 1970-1980." make the Washington D.C. one of potential value to understand its geography and its politics (respectively, and again, there might be more specific links that would be better targets). No one is advocating making the major geographical features as orphans, but just to avoid linking to it just because it's mentioned. --MASEM (t) 14:06, 1 November 2012 (UTC)

• Agree with Masem. And JonRichfield, I spent 10 seconds flicking through your gigantic posts (and those of one or two others). If you keep it to a shortish paragraph or two at a time, there's more chance your fellow editors will actually read it. I just don't have that kind of time and energy. Tony (talk) 21:31, 31 October 2012 (UTC)

A lot of the comments here involve the same assertions, frequently rebutted and/or yet never properly explained, and the same unanswered questions that have swum round this issue for years. For example:

• the idea that links are only there to explain what something is at a basic dictionary level; even though WP pages are in fact encyclopedia (of a sort) entries, with a mass of detail, not simple dictionary definitions. If we think people, or the "average" person, already know what X is, why have an entry on it at all?
• the distinction between so-called "high value" and "low value" links. Who determines this, on what criteria? I don't know what it means, especially not in any objective sense. Is every link "high" or "low" value in more or less the same way for every reader, regardless of who they are, where they are from and what they are looking for here?
• the idea that we can set a bar as to what the "average" reader "knows". Where, in fact, is that bar?
• the "dilution effect". What is this and what is being diluted? What effect does it have and what evidence do we have for that? Does it even matter if it does exist?
• the "distraction/diversion effect". Similarly, what is this and what are people being distracted or diverted from? What effect does it have and what evidence do we have for that? Does it even matter if it does exist?
• combining the above two points, how are "low value" links at the same time "distracting" and "tempting targets" but also likely to put readers off following them?
• why, at the end of the day, we cannot credit people reading WP pages with enough wit and intelligence to exercise a rational choice as to which links they use, rather than rather patronisingly asserting that they will somehow be "tricked" into clicking onto a "useless" or even "bad" link?

Some readers will want to follow some links and not others, others will follow different links. The same reader may choose to follow a different link on a different day from the one they might have followed the day before. Readers will generally not follow links to things they have no wish to follow up. Why is this a problem and why is there a small group that arrogates the right to police that process and excessively limit the reasonable options that might be available to readers? WP may be the encyclopedia anyone can edit, but some of us seem to think it very definitely should not be the encyclopedia that anyone can navigate around with relative ease in the way they might wish to. N-HH talk/edits 23:22, 31 October 2012 (UTC)

The way to consider a good linking approach on WP is that we are trying to make a good information map - in that when I make a link from A to B, there should a reasonably way of getting from B to A via one or two additional pages in a logical sense (ideally we'd love any link to B to be reciprocated with a link back from B, but that's really hard to work towards considering the breadth of some topics). When such links are possible those are high value links - they form the framework of this information map to make it much easier to navigate closely related topics. Low value links are one where there's no logical path to get back to A from B. (One cavaet: there is the factor of definition links of uncommon terms and names which are necessary to include because, well, they're uncommon unlike the rest of the English language. I would not call these low value.)

A well-established information map makes it much easier for researches to get around both through link navigation - knowing that every link will take them to some highly relevant topic - and through the meta-nature of the information map being able to identify core topics of an area. Low value links muddy that map, and devalue the usefulness of Wikipedia to the researcher. Of course low value links make us look like TV Tropes, but we don't want the WP's reader's experience to be similar to there.

One thing we have to remember is that there's always a call "Won't someone think of the readers?" If we succumbed to what the readers wanted, we'd have endless fan guides of popular TV shows, coverage of every sporting event, every garage band ever, and the link. We're trying to build a work that will last the test of time, and to that end we have to take a more demanding approach to how we write it, including how we build out the linking. It may be offputting to some readers, but it doesn't seem to be a significant majority of the readership. --MASEM (t) 23:54, 31 October 2012 (UTC)

Your "The way to consider a good linking approach on WP is that we are trying to make a good information map - in that when I make a link from A to B, there should a reasonably way of getting from B to A via one or two additional pages in a logical sense (ideally we'd love any link to B to be reciprocated with a link back from B, but that's really hard to work towards considering the breadth of some topics)" is a downright embarrassment. Did you read it before posting? Some links should indeed be reciprocal, but if you will excuse a 4-letter word in open forum, there is no general logical reason why all of them should be, even in a perfect world; in fact, in a perfect world major classes of links would not be reciprocal. Consider an RFC I received recently for the article Urine. Someone wanted to link to the article Urolagnia because urolagnia intimately concerns urine. I (wait for it…!) opposed such a link because it had nothing substantial to add to the study of urine. I did however suggest the idea of a link from urolagnia to urine because, though urolagnia is not fundamental to the concept of urine, urine is fundamental to the concept of urolagnia. Note well however, I did in my comment remark that if anyone could think of any worthwhile text on the subject of urolagnia (relevant, encyclopaedic, NPOV, and all that) to add to the urine article, then I would waive my objection. I am not sure whether one-way or reciprocal (logically valid, constructive) links predominate, and frankly I don't much care. You will note that the justification or requirement is contingent. Today's "no-no" might be tomorrow's "definitely".

You go further: "When such links are possible those are high value links - they form the framework of this information map to make it much easier to navigate closely related topics. Low value links are one where there's no logical path to get back to A from B. (One cavaet: there is the factor of definition links of uncommon terms and names which are necessary to include because, well, they're uncommon unlike the rest of the English language. I would not call these low value.)" I almost blush in asking you to re-read and consider what you said! What you need is a good dose of graph theory 101. Just consider the value of your map when all the most prominent nodes have the lowest proportion of links from articles that most intimately are concerned with them. You could wind up with larger numbers of links directed to minor subjects than to major topics. You gotta be joking! Have you at any time and place actually formulated any sort of research programme based on these ideas? The mind reels...

You also say in part: "A well-established information map makes it much easier for researches to get around both through link navigation - knowing that every link will take them to some highly relevant topic - and through the meta-nature of the information map being able to identify core topics of an area. Low value links muddy that map, and devalue the usefulness of Wikipedia to the researcher." This is largely inconsistent with the very nature of the links you have been describing in previous paragraphs, including in the very same post. What is more, these mystical and unspecified researches are not what the links are for. They are for the benefit of the readers, not the notional researchers. If you want a researcher's data map, by all means find or propose some other tool, some more valid tool, for them. Their needs, as nearly as you seem to be trying to characterise them, patently cannot be the needs of the readers. And the links in their current form are not in themselves the right tool for your mapping. I already said so, didn't I? What part of that did you fail to understand? Hint: Ask yourself with as little handwaving as possible, i.e. as if you were setting up a research proposal, how you would go about using the current links, as used by readers and editors, in a material way.

I think you have said something of this type before: 'One thing we have to remember is that there's always a call "Won't someone think of the readers?"' This surely cannot be true? What a strange concept! Why should anyone care about... errr... Hm. Frankly Masem, I think you have your knickers in a twist. What you are attributing to the desires of the readers, actually stems from the desires of a lot of writers. What percentage of any of the garage bands get more hits than major topics? And why should mainstream users and editors care... much? If we succumbed to what you want, we'd have a work that would be doing well to last as long as the Edsel. The demanding approach to how we permit people to write it is the only weapon we have against the groupies and trivialisers, but that has little to do with linking. We have your expert opinion that it (whatever your antecedent to it might be) may be offputting to some readers, but, you explain, it doesn't seem to be a significant majority of the readership. Masem!!!! It doesn't have to be a majority! A large minority would be enough, if we were counting concerns along with counting noses, let alone heads. JonRichfield (talk) 15:25, 1 November 2012 (UTC)

First, as Tony said, you need to learn to shorten your replies. This is way too much to read.

But to the points: I'm not saying that when you make a link from A to B, that there will always be an immediate link back from B to A, but it may be on a closely connected page to B. For example, linking from a film's page to an actor that starred in it may not have the film immediately linked on the actor's page if we employ a separate filmography article. In your example, I am surprised that you aren't discussing Urolagnia under "Other Uses" of urine because, despite how gross or whatever that may be, it's a highly related topic.

Gross doesn't come into it; the point is whether anyone has anything of value to say about it. I did tell you: " Note well however, I did in my comment remark that if anyone could think of any worthwhile text on the subject of urolagnia (relevant, encyclopaedic, NPOV, and all that) to add to the urine article, then I would waive my objection." Still holds. No one has taken anyone up on that one; feel fully welcome as far as I am concerned; My religious commitment in such matters is that we serve the presentation of encyclopaedic information content to those who want it, gross or not. And as for your indirect routes back, yes, I had understood that point even before you raised it, and I was bearing it in mind. No difference. JonRichfield (talk) 12:03, 2 November 2012 (UTC)

The other factor is that there will be proper name definition links that we don't immediately expect a return link. While we avoid linking major geographical features, identifying "Smalltown USA" via a link is appropriate regardless of context where it is used.

Yeah, I more or less got that, but I thought I explained that I don't see how it helps your map that Smalltown, Liechtenstein gets linked and New York, NY does not. I would have thought that the relative frequency of links would be a valid datum in research, even without having to brandish a whip over the heads of linkers. If natural linking brings a landslide over your head when you click on "what links", then the tool your researchers lack is an aid for sorting and structuring, not trying to impose a uniform behaviour among both on assorted readers and editors. Nor do you make it plain how you see that fact as being consistent with whatever your definition of link quality or research might be. JonRichfield (talk) 12:03, 2 November 2012 (UTC)

Even if we consider what the bulk of editors want, we don't go that direction and instead use consensus built on a solid principles to build a professional encyclopedia with permanence. I've butted heads with Tony and others that are responsible for building most of these MOS pages, but I do agree that they are bringing proper professional writing concepts to an open wiki where there's a broad swath of editing styles otherwise at play, and to help normalize them to something reasonably consistent that better serves the work's academic purpose. There are definitely elements of the MOS that likely are only favored by a minority of editors but they have reason and practice behind them to show them to be effective aspects of building a good encyclopedia, and reduced linking is one of those elements. It's an art, though, not a science, and there's room for discussion on where lines are drawn, when a term is common or not. But major geographical features is definitely one of those that should be rather straight forward to avoid in prose, given how frequent such words are used alongside common nouns and the like. --MASEM (t) 22:39, 1 November 2012 (UTC)

If the specific items that you favour in the foregoing paragraph are "proper professional writing concepts", then proper professional writing concepts are not what we want in an encyclopaedia. Since I regard consideration of function as one fundamental aspect of proper professional writing concepts (and also of research!) you place me in an embarrassing position, given my pathological unwillingness to express discourteous implications. The question of linking, as I have said and said, is not one of prettiness of style, but one of function. Prettiness is a matter of taste, and a matter that can be served in various ways, including some that I have mentioned (want me to explain? Again?) Function is a matter of the rationale for WP and its role in the service of humanity. Get that right and then come back and tell me all about proper professional writing concepts! As for relying on building "consensus", for that you need some persuasive logic, not handwaving. Try matching my list of link functions and possibly expanding it for a start. Try defining the nature of at least the research aids you long for, never mind the research. Consensus requires work on both sides you know. JonRichfield (talk) 12:03, 2 November 2012 (UTC)

• It's quite enough that a reader knows of Washington D.C., and quite trivial that [he] doesn't know much about it in the context of the subject's biography. The important thing is what the collective linking brings to the reader about the subject of the article. There's just enough patronising and facetious verbosity above why we should plaster ever more links to give the reader that ever elusive choice that "vicious unlinkers" are depriving the world of, and little about how knowing about Washington DC – above that it's the capital of the US located on the east coast – actually deepens the readers' understanding of the subject. Sure, it's the fellow's birthplace, but the qualitative increment to knowledge of the subject it brings is zilch. The same can be said of the need to link common garden words (comedian, television host, actor, writer, and scientist – in quick succession) that do little to deepen that understanding. --218.103.34.213 (talk) 01:49, 1 November 2012 (UTC)

The fact that you do not, outside the context of what a given reader wants, know why he should not see why he should benefit from a link, is not of interest. I refer you to the previously compiled list of uses for links and invite you to extend it. A decent link will firstly enable a user to see a lede or similar item "Ooohhh... that thing... Right." and maybe elect to go deeply into the matter. "Hey, what's all this about; I had no idea..." His call mate, not yours, Not Masem's, not mine even! JonRichfield (talk) 15:25, 1 November 2012 (UTC)

I don't disagree that there needs to be some discretion and editorial judgment about linking, and that there needs to be focus (with the key being placing and keeping links to related or relevant proper terms rather than being based on guesstimates about what the "average" reader supposedly already knows or might want to know; I've also always been quite happy to see common generic words/terms unlinked for example, other than in articles about related generic words/terms). The issue of course is what kind of judgment is made and where the axe falls. I don't agree, however, that this can be done on the basis of the kind of assertions often made in these debates, which purport to be objective, quasi-scientific analysis, but which are nothing of the sort; or on the mass-stripping of links to so-called "common terms" such as "France", pretty much regardless of context. Many of those assertions – along with the above comment about what is "enough" for a reader to know – are incredibly arrogant and basically saying "if I don't want a link, or don't see why anyone else might, no one else shall have even the option". Who appointed 218.103 ... to decide what other readers want and how they might wish to navigate between pages in an online reference source? At the same time, there's nothing special about my view on linking – if someone disagrees with me as well as the more radical delinking, they're perfectly entitled to argue their case. They're not wrong, which is so often the way this debate is conducted, they simply have a different opinion. N-HH talk/edits 08:51, 1 November 2012 (UTC)

I may have missed this answer in the wall of text, but are we in general agreement about linking Washington and Los Angeles in the infobox? If so, is there an admin who cares to do that? The page has been protected. InedibleHulk (talk) 04:22, 2 November 2012 (UTC)

We used to have people going around linking every single word in infoboxes, as though the aesthetics of wall-to-wall blue were the key. I don't mind the idea of a single, uniform colour—though I'd not choose that garish blue they have as a default—but the problem is that readers won't understand that they are links, then. And it removes our ability to point to the most important, useful items as links, too. Tony (talk) 06:18, 2 November 2012 (UTC)

Tony, the reason we have difficulties with such points is very simply that some folks don't understand the proper functions of links. If every link is functional, then it doesn't matter whether there is a lot of blue text, sometimes even solid blue in the likes of infoboxes or taxon lists say. Anyone so stupid as not to be able to work out in a reasonable time span which is which and what for, isn't going to benefit from the links anyway, nor even from unlinked text. Masem was explaining about our superior English Education System and how WP wasn't for the lesser breeds, right? Well, what kind of high-school product would you expect to be unable to work out that if one blue word could act as a link, possibly another blue word...? And if a word weren't blue, why clicking on it wouldn't do much linking for anyone? I have known unlettered farm labourers who would pick up that sort of thing over a cup of coffee; where do you and Masem fit into that scale of intellect if you couldn't cut it, or believe that anyone else could? JonRichfield (talk) 12:03, 2 November 2012 (UTC)

... or even unfettered farm labourers. Now to be serious, I support linking wherever it is reasonable: one does not have to click on it just because it is there. But it just might help a reader, so why not have it there?

-- Gareth Griffith-Jones/The Welsh Buzzard 12:11, 2 November 2012 (UTC)

Right you are Gareth, but look you, that is what we have been telling them for years (feels like quite a lot of years now, but there haven't been computers for that long) and we have even explained what a simple matter it would be to overcome most of their woes with a simple option, but the penny will not drop, and fettered to their preconceptions they remain... JonRichfield (talk) 12:52, 2 November 2012 (UTC)

Linking without consideration for relevancy would make us more like TV Tropes, where readers get lost in jumping between page after page with minimal logical connections. There, that behavior is okay because the site is being used for entertainment purposes, but here, on WP, we are to be a research work. Links in prose that are not well chosen are distracting and can lead the reader off the purpose of reading the prose, defeating the purpose. Some may come to WP for that type of experience, but that's not the goal of WP.

Also, as soon as you start saying "Oh, we shouldn't worry about linking too much, our readers know what to skip", then the only logical conclusion is to link nearly every word in prose, since all of those are potential targets. A slippery slope, perhaps? Maybe, but there's a very grey line between sparse linking but without considering values and linking every term - there's no place to say when to stop linking. But it is much easier to define where to stop with high-value germane links before they start to become low value, so it's a better method of selection that avoids edit wars that link every single term. --MASEM (t) 13:02, 2 November 2012 (UTC)

I see no problem linking in the infobox the first time the terms used there, only because that's not prose. --MASEM (t) 13:02, 2 November 2012 (UTC)

Gosh Masem, what would we lesser mortals do without your guidance? We would have Welsh bards inveigling us into all sorts of excesses that we had never thought to eschew! Writing or linking without relevance for example! Did you notice with what devilish cunning he totally omitted to mention relevance? How could we have resisted his Morgana le Fay-like spells? We would have succumbed and gone on linking all sorts of words whether necessary or not! Good grief; he's got me doing it now and even repeating some of those whether necessary or not! Doomed we would have been, is it? Looking glass logic isn't bad enough, we have to deal with Merlin's spells as well? But never mind; bring on the buzzards. Masem has the necessary invocations: he knows the the goal of WP, research work only has high quality links and links are distracting (how could we ever stop ignorant people from clicking on them, no matter how often they have done so before?) And your generosity does not end there. You provide novel bogies to scare off evil spirits that may have begun to develop resistance to the old ones; I wonder how long they will last: ...can lead the reader off the purpose of reading the prose, defeating the purpose? The sheer unexpectedness of such dire certainty has a lofty tone calculated to tempt the veriest Welsh choir into parts and chorales, before the rest of us had even braced ourselves for such a thing, imagining in our innocence that the average reader would realise, as Gareth in his evil cunning put the thin end of the wedge in temptation's way, that: "...one does not have to click on it just because it is there". Aah, the poor silly lambs to the slaughter! You rescued them in the nick, didn't you: '...as soon as you start saying "... our readers know what to skip", then the only logical conclusion is to link nearly every word in prose.' There! What did I say? That Welshman almost had me going! Quick, repeat that logical conclusion, logical conclusion invocation a few times! And Masem, don't you go away now, we need you to explain how much easier it is to define where to stop with high-value germane links before they start to become low value. We hadn't realised how easy it is. Some of us were gaily trying to place links where we thought they were likeliest to be helpful, instead of counting the beans (one link for this article, one for that infobox...) This is a difficult matter with subtle difficulties such as "...linking in the infobox the first time the terms used there, only because that's not prose". You must please explain why we should not link in prose the first time; this is a deep problem. And what when every word in the infobox is a technical term that only occurs once? Quick, before I have Gareth tempting me into one of those edit wars that link every single term! JonRichfield (talk) 16:37, 2 November 2012 (UTC)

There's no use in replying if you are going to be that condescending. --MASEM (t) 16:56, 2 November 2012 (UTC)

Different cultures have different expectations of humour ... two well-known stereotypes in Britain are that Americans don't understand irony and that Germans have no sense of humour. Whether these statements have any validity has been discussed in a BBC News article. -- Gareth Griffith-Jones/The Welsh Buzzard 17:29, 2 November 2012 (UTC)

Condescending, hm? Masem, that is a convenient but dangerous word. Like democracy and racism it means just what Humpty Dumpty pays it to mean. Condescending? How many of the quotes or near-quotes of your own words in my foregoing paragraphs were distilled condescension? How many were repetitions of words of your own or other linkage control-freaks, if not actual de-linkers? Your dismissals of views or proposals that you did not personally accord the dignity of rebuttal, simply because you were incapable of coming up with anything better than repeating that they were the slippery slope to all-blue articles; how many of those dismissals were anything but condescension without the extenuating merit of validity?

Now, Masem, I am not picking on you in particular; you are not the worst specimen by any means (no names, no pack drill) but where I come from we have a proverb: "As jy jou vir tameletjie uitgee, moet jy verwag dat mense aan jou sal kou." You might find that in controversy it makes for a more peaceful life if you are careful in attributing attributes, particularly attributes attributable by reverse link; the high-quality linked attributes, one might say.

Mind you, there might well be a more insidious culture clash here than either of us has been allowing for. In this connection I see that Gareth has appended a remark that you might well take to heart... or to head...

But that still doesn't mean that there is any use replying to me before you have the necessary substantial and logical material for cogency, and have it suitably marshalled. We need something better than the hysterical subjunctive for a change. JonRichfield (talk) 19:04, 2 November 2012 (UTC)

I have never said anything condesending in this discussion nor have ignored the other views. I simply recognize that the other views make creating a professional work unobtainable. On the other hand statements like "Gosh Masem, what would we lesser mortals do without your guidance?" or "Looking glass logic isn't bad enough, we have to deal with Merlin's spells as well?" have absolutely no place in debates on WP. --MASEM (t) 19:19, 2 November 2012 (UTC)

Even if making fun of Masem is humorous, it is also pushing the envelope of Wikipedia:Civility. And it doesn't seem to contribute to the discussion; why make fun of removing low-value links, when everyone seems to agree that the is the extreme example of a low-value link we should remove? The substantive issue is how low is low, not whether we should unlink anything at all. The "who are we to decide" argument similarly fails, because we all agree we are to decide on not linking everything. Art LaPella (talk) 20:26, 2 November 2012 (UTC)

I'm still not sure who "we" are or would be. Anyway, in response to something a while above, can I simply say that I'd happily see more/a focus on links in infoboxes, often as an alternative? Not least because it would hopefully satisfy all sides in this and allow us to avoid "excessive" linking in running text while offering the option of links being there for some people, even to "well-known" terms"? N-HH talk/edits 23:08, 2 November 2012 (UTC)

Let's not get too deeply into incivility or counter-incivility. I apologise to Masem and to any readers who took offence at the things I said or the way I said them. What I cannot apologise for is the substance of what I said, nor at taking offence myself (not specifically at Masem's stuff; he is no worse than some. Disagree with his views and intentions though I may, he at least takes the trouble to express them and has important considerations in mind. Which is more than I can say for some others.)

But I in turn am not here for kudos for my deathless (and largely unlinked) prose. In no particular order, I have variously questioned the cogency and rationality of the way some parties urge and apply their favoured forms of link control, and more particularly questioned their reasons, justification and alternatives, when they mooted any; I have proposed cheap and simple means for eliminating at least one of the possibly material objections to excessive, or perceived excessive, linkage, means that would entail neither loss of function nor increased operational burden; I have rebutted repeated slurs, amounting to trolling, concerning my dishonestly attributed desire for every text word being blue, or even for linkage irrelevant to the theme of the article; I have pointed out the irrelevance of the mechanism of linkage to some of the functions proposed as justifications for control of links a la the personal tastes of certain people, and have urged that as far as practical the linkage and its manifestations should be under the control of the user, whose needs are after all the first and ultimate rationale for WP; I have listed roughly a dozen criteria for when and how to apply links; and all this has gone on intermittently for months in an exchange that was old when I entered it, and in which the predominant response of the forum establishment was offensive and dishonest attribution of motives and illogic, justified by the likes of the observation that some French articles had looked uncomfortably blue to some readers (not having seen the articles in question, I cannot comment on them, but since no one was recommending anything of the type, it is hard to imagine their relevance). Masem, to do him justice, claimed that "...links in prose that are not well chosen are distracting and can lead the reader off the purpose of reading the prose, defeating the purpose..." and that certain classes of links would sabotage aspects of WP necessary for a certain unspecified form of research ("...that the other views make creating a professional work unobtainable"). About the rationality of the former point, least said soonest mended, but the second one he still has not particularised, and he hasn't even hinted at why the research in question is more important than the interests of the user, whether links designed to serve the user are the right tool for his research, or ever could be, or even whether, assuming that the research is in any way justified, some other mechanism might not be better calculated to serve it.

Bottom line: sister pot is black. Will anyone in a position to do something about it, kindly assist in burnishing both her and brother kettle? JonRichfield (talk) 09:18, 3 November 2012 (UTC)

Below the bottom line

Amen to that. Well said JonRichfield!

Just to be clear that I meant to cause no offence to anyone by my entering this debate so near to the close (I imagine it is near the close, isn't it? – "there's Welsh for you!") and writing in the style that may have appeared flippant. Please remember that a strong theme of sarcasm and self-deprecation, often with deadpan delivery, runs throughout English and Welsh humour. Emotion is often buried under humour in a way that seems insensitive to other cultures.

Kind regards to all involved here, -- Gareth Griffith-Jones/The Welsh Buzzard 10:05, 3 November 2012 (UTC)

We are a reference work, first an foremost, and that means ours style of writing and presentation should best serve those using it as a research tool rather than those that are reading it casually. When we can serve the casual reader too without hard to the researcher, sure (eg: links in infobox would be an example), but the use as a research tool has to take precedence where there's conflict.

Also you may have suggested points before but as Tony indicated, your posts are simply too long and they are being ignored. There is a time and place for verbosity and fluidity with languages - talk pages on WP are not one of them. You may be better able to state your point with shorter, direct statements. --MASEM (t) 11:01, 3 November 2012 (UTC)

You don't seem to remember Tony's full post: "Agree with Masem. And JonRichfield, I spent 10 seconds flicking through your gigantic posts (and those of one or two others). If you keep it to a shortish paragraph or two at a time, there's more chance your fellow editors will actually read it. I just don't have that kind of time and energy. Tony (talk) 9:31 pm, 31 October 2012, last Wednesday."

A case of the kettle calling ... et cetera, don't you think?

-- Gareth Griffith-Jones/The Welsh Buzzard 11:18, 3 November 2012 (UTC)

Masem, you still fail to make your idea of research and reference tools clear, to such a degree that I find it hard to believe that you have any clear idea of them at all. You even fail to make it clear why casual users should be regarded as lesser citizens, or how their activities are fundamentally less important. For a start, if you are thinking of research into subjects such as the structure, usage and data mapping of WP, then what on Earth makes you think that non-trivial control of editors' linking and cross-linking of texts by the whims of other editors with conflicting ideas will provide unbiassed, objective material for research, let alone a data map? Please help me (I have been trying hard enough long enough surely?) by outlining some sort of research programme, with objective and protocol (you need not trouble yourself with anything arcane and watertight, just a thumbnail sketch of a conceived line of work) that you would base on such a WP and that would not work on a more freely-linked, even over-linked WP. If you did that, we might be able to stop tediously repeating assertions at each other.

If OTOH you are thinking of "research" in the sense of using a reference work as a source of content, then I am stunned at the idea of a user who has the mental equipment to engage in anything resembling competent research, and yet cannot effectively read text that contains the occasional word in blue, even if it might more occasionally appear in blue in more places than one, or next to another word in blue, or appear in a context where some other readers might not want or need a word in blue. In either context you certainly don't clarify your idea of the differences and conflicts between the rights or the value of casual and "research" users. And most certainly you don't make it clear why what you conceive to be casual users should want any more, or fewer, links than what you conceive to be research users. I invite you to inspect the list of suggested reasons for linking and ticking them off against casual and research users, and against justifications for various densities (ugh!) of links.

Masem, there is very, very, little conflict between the most functional and most readable linking in an article and far greater variation between the titre of unsatisfactoriness to different readers, of differing styles of text linkage, or even the same reader on different days; such things are subjective and volatile, and as I have said, usually trivial; what comes closest to utility and stability is linkage that enables the reader to read what he chooses and to follow data references and relationships as he chooses and as smoothly as possible, without having to jump around looking for links and then having to type them in with one finger, and then having to try again and again in case the link was not under the name or spelling or even heading that he had guessed, and finally having to retrieve the spot where he had left off. No doubt I am overlooking a vast and silent majority who would find such a process to be much less distracting than perhaps ignoring or sometimes clicking on an occasional blue word. (Just a day ago I had to create a link for spigot which had been originally entered as spigot. A lot of people had apparently passed by the link without reaction. It took me quite a while to decide what to do. As a user of the article, whether "research" or "casual", what would you have appreciated most? A wrong link, or no link, or a functional link? And in which part of which article? I still am in some doubt whether to create a new article for spigot, though I am in no doubt that if I were to, I would link to it. Let me know if you have rooted views on the subject, will you?) As I have said before in effect, if A creates text and includes a link (say to strangeness) and B regards it as ridiculous, not to say strange, to link to anything so obvious and edits the link out, and A accepts, there is no problem; out it goes. But if A says: "Hey, do you mind? I deliberately linked there", then unless B has truly crushing reasons, the link should stand if only a matter of taste divides the options. In fact, even if B had been the original author, then A's strong preference for inserting a link still should rationally take preference. No one has the right to erase from a public map, everything that he is familiar with or thinks he is familiar with. Same with data maps.

Masem, just a little while back someone (can't remember who) castigated someone else for lack of respect for his fellow members of this forum. Someone else astutely pointed out that culture shock or culture gaps might account for a lot of the perceived offence. Conversely, I think that even such a peacemaker might find difficulty in rationalising the insult to one's fellow members, of implying that people who surely must have read a textbook once or twice, are likely to be unable to find a list of points in a few long paragraphs of casual, unedited text. I certainly could not bring myself to insult them thus, but once again in case I have misunderstood your reasons for not replying to the points I made, I invite you to inspect the list of suggested reasons for linking. You then would perfom a signal service by demonstrating their merits or lack of justification or feasibility, and if that strikes you as too much PT, let me know and I'll assist in fishing them out for you. (Really! No sarcasm, no irony.) (BTW, on my screen this fits onto one page; I hope that is satisfactory. If not, please let me know which points I should have omitted.) JonRichfield (talk) 15:16, 3 November 2012 (UTC)

Your spigot example proofs the point of only linking high value links. When you overlink and include low-value links among high-value ones, a reader is going to have no idea what relevance the link is to the article and thus will less than likely click away to learn more if the average "value" of the linked article is low. So in the case where an article linking both low- and high-value links, and the word "spigot" (which is a common English word) was linked to the article on the organ, not to the physical device, I would expect few would click on it. But if we only had high-value links, and "spigot" was still linked, I would expect readers, knowing that all links are high-value, to want to follow that and will learn how that term is used relative to spider anatomy (something I just learned now). In other words, yes, "spigot" should be linked to the organ article irregardless, but the chances of readers following it drop if it is surrounded by low-value links as well. We already have similar advice in the MOS: we don't use "surprise" links where one term is in the pipe part of the link but the actual link is a very different term. That's extremely diluting, and makes the reader have less confidence in the value of links. --MASEM (t) 15:26, 3 November 2012 (UTC)

Thanks for a courteous and practical reply. (Really! And I am not being condescending either.) However, it is not fully persuasive (or of course, we would not have been at odds in the first place). I in turn point out that I prefer to link wherever the link is perceived to be of value to some bona fide reader of the text, and consider that when someone reads text such as "spigot" or "barn" or "strangeness" or "head" in a context that makes no apparent sense, he is likely to notice that it is linked when he does his double-take, even if it is in a string of blue. And what do we do if we notice a link where text makes no apparent sense, and that text is linked? We click, don't we? I certainly do. I accept that we agree, essentially fully, when it comes to linking an isolated, unfamiliar technical term, such as pseudanthium or osmeterium and also when it comes to linking the term the first time in the article. I suspect that we enter less amenable territory when we deal with the likes of lists of species names or other taxa that might not occur elsewhere in the article at all. Such a list would almost infallibly be solid blue, but I have an idea that you would tolerate that. If not, then I am afraid I have no idea what you would tolerate instead. If I am correct, we disagree in that I also would almost certainly link the first occurrence of any of those items in the open text (prose, if you like) as well. Where I think we are very likely to disagree, is that I also would be very likely to link the first occurrence of the term in each section, and at least once on each page on which it occurs. Feel welcome to correct me where I have misinterpreted your preferences and statements.

When it comes to unambiguous use of terms that many of us would regard as too familiar to be worth linking (though long and bitter experience has taught me to be very, very distrustful of such familiarity) then as I see it, what matters is the sense and structure of the sentence and its place in the article. (Surprise, surprise!) For example, we were diverging on the subject of linking Washington D.C. No doubt the impression that you got was that I would link Washington willy-nilly, and continue linking it wherever it occurred in the article. Well, actually, I don't link willy-nilly as a rule, but try to link where I see it as being useful, or reasonably likely to be useful, to someone, sometime, maybe... Consider a sentence in the middle of some article that runs: "between the 1950s and 1980s he spent his life in major cities such as Moscow, Tokyo, New York, Washington, and Canberra..." Unless there were some particular circumstance that does not occur to me as I write this, I would not link any of those city names. However, if some other editor came by and linked the lot (assuming that they had not all been linked in immediately preceding text) I would not revert his linking them. He might well have a better insight into what readers need then I do. However, I would link: "He made a special study of the demographics of cities such as Washington and Canberra" and I would be very bloody-minded about any mindless unlinking even if the unlinker could demonstrate that he in particular could run the linked info off pat.

BTW, quite adventitiously, I just now had reason to create the following phrase: "The distal tagma of an arthropod leg..." How do you feel about the links? (None duplicated on the same page.) High or low value? And in either case, higher to biologists, or to poets, or to foreign readers, or to school children? What parts would you change? How would you edit it to avoid using such horrible language? JonRichfield (talk) 18:48, 3 November 2012 (UTC)

• To avoid easter egg links, you could write something like " "He made a special study of Washington's and Canberra's demographics", wording the subset "such as" differently; readers are unlikely to click on plain pipes to well-known cities. "The distal tagma of an arthropod leg..." looks good. Tony (talk) 00:19, 4 November 2012 (UTC)

Thanks Tony. What you had to say sent me off for another look at easter eggs, which in fact I had forgotten. I accept your point and your rewording, which is in fact in line with my intention, namely to proffer an option to link usefully to relevant info, even if the keyword suggests a stunningly obvious subject. The tagma thing I am glad we both are comfortable with, but I did want to highlight the fact that in technical articles it often is possible to have very high densities of blue words where each link is none the less individually justified. Cheers, JonRichfield (talk) 11:54, 4 November 2012 (UTC)

Jon, yes, that was not an easy one to reword, either; usually easy to work the more specific into the grammar. Tony (talk) 13:56, 4 November 2012 (UTC)

I agree with the scientific term example that there's little else one can do to reduce linking; they are all uncommon words and should be linked to start.

To go back to the "spigot" example, again, you have to envision the case where a reader has become familiar with a version of WP where too many low-value links exist throughout articles, that it is impossible to tell easily - without clicking (eg in context) - if a linked term is immediately relevant or not to the article; eg where very common English words are linked just because they can be. In this specific case, the reader seeing "spigot" linked in the case above will likely ignore the link because it still is impossible to know whether it is a real intended link, or just the poor quality link that goes to the article about water taps. When links are rare, the fact that "spigot" is linked should make the reader wonder why a common English word is linked, and thus follow it to learn what spigots are in relation to spider anatomy.

And that brings this around full circle - we have to agree that we need to assume some lowest common denominator of language and other knowledge for building en.wiki articles - as to know what terms are best to link. Anything that reasonably falls out of that should be linked (the tabia example) as definition links for most readers, but when we talk about common words we need to be much more careful. And that point that we have selected is a standard high-school level of English education - and it is hard to imagine that a student would not be taught basic world geography during those 9-12 years. This is why we avoid linking to common names in geography when they aren't relevant - just like linking to common words without careful consideration, readers will be less likely to follow links, defeating the purpose of the linking system. --MASEM (t) 15:34, 4 November 2012 (UTC)

But Masem, it is not easy to tell where a full circle swallows its tail, if you will excuse the reference to Ouroboros. We generally have the responsibility to decide what information to include as a duty to our users, and what to omit as a matter of common sense. As Tony has just demonstrated with my easter eggs, it also is a matter of how we present it to make it useful to (sometimes advanced as well as naive) users (we are all naive outside our own spheres of knowledge). Now, if some people find it irritating to read material or see links to material that they regard as obvious, while others had no idea of it (I am perfectly certain that just as you didn't know of arachnid spigots, you could find hundreds of simple examples which would be obvious to you, but blackly opaque to me until I click on their links), then as I see it, giving the option to the naive or young user trumps conceivable irritation to the advanced user, even if we ignore cases where it has little to do with ignorance, but with choice of what to digress into and when. (The list of reasons for lining.) And as I keep nagging, if anyone really wants to split skulls after reading his nth link on the same page, why shouldn't he be able to click on a toggle instead of uselessly screaming at the screen some equivalent of "SHUT UP YOU !@#$%^&!!!", and still having the capability of clicking on unhighlighted links when he pleases. Frankly, if I had such a toggle at my disposal, it would take a really pathological example of overlinking to make me bother to click on it, but it seems some people would love it. And that is even if no one comes up with a much smarter option.

In short, what I am insisting is that there is no guaranteed correct option, but that there is a clear direction in which to err for preference, the lesser of evils in fact, plus cheap possible means for partial alleviation. JonRichfield (talk) 20:35, 4 November 2012 (UTC)

But we don't cater to young readers or those unfamiliar with the language. If we did, most of our technical articles would need to be rewritten to start at a much lower level of understanding. Instead, the Simple English wiki was created. When this principle is applied to linking, the line for when to include low-value links is very fuzzy, and wants to trend towards linking anything that can be when groupthink of editing is applied. The use of keeping links to those of highest value is a much easier line to define - there's some fuzziness to it but it does not trend towards an extreme. It also makes us look much more professional to our target audience as it avoids de-valuing the entire linking system. This again is why we have set an expected education level that we expect our reader to have, so that we don't have to worry about how to scale back to meet those outside of it. I've watched the overall Wiki mature over the last several years and its clear we are heading towards being a professional research tool as opposed to a causal site for reading (introducing concepts like notability, verifyability, BLP, citation improvement), and our linking system should reflect that. Yes, if there was no impact on the profession research use of including more links, great we should do that, but there is definitely a negative impact and thus we avoid doing that. --MASEM (t) 21:37, 4 November 2012 (UTC)

@Masem. I deny your flat remark that WP doesn't cater for young readers or those with language difficulties. Certainly there are limits to what one can do. Certainly some of our topics simply do not lend themselves to really simple treatments, no matter what anyone told Rutherford about the man on the Clapham omnibus. Of course, subjects that really are heavy going, and other subjects that look much simpler, but actually are full of intellectual traps, commonly benefit from particularly high frequencies of links, always depending to some extent on how finely they are split into separate topics, but there are very accessible subjects that benefit equally, and the more the constructively applied links, the higher the didactic value is likely to be. I should have thought that obvious, but if it is not, do give it a bit more thought. This said however, there is a real slippery slope risk here. In our discussions you and I already have encountered the fact that when one deals with a specialised subject, nonprofessionals in the field may have difficulties with what insensitive professionals regard as obviosities. Dealing with such problems, as I hope you realise by now in the light of earlier exchanges, is not something I represent as the only, nor probably the primary, reason for using links. However, links certainly are important tools for dealing with this class of problems, and in fact for dealing with them at many levels and in many ways. I even have been inveigled into writing or editing one or two (can't remember) simple Wikipedia articles. My efforts were accepted, but I found the work personally gruelling and unsatisfying, so I concentrate on material where I find that I can deal with topics at a more natural level. Now, this slippery slope that I mentioned is the difficulty of deciding firstly, the level at which one decides to condescend (a word that does not mean at all the same as "patronise", please note) to the reader who needs help. Personally I propose that any reader who never needs help at all won't need Wikipedia at all, and also that such a person would be fictitious. Also personally, I reckon that any assistance that can functionally and courteously be offered to any user above kindergarten and non-Anglophone Chukchi level should indeed be available (not intending any discourtesy to the Chukchi speakers of course, but the linguistic overlap with English is so slight that we could hardly make allowances for it in planning our articles. (Interestingly enough, in my contact with gifted primary school children, I found their receptiveness to quite advanced concepts quite startling, so be careful what you say; you might get egg on face!) Within such limits we should bend over backwards to accommodate the needs of whomever we can, and that is what should define our "target audience", not our own tastes and opinions of what we think readers should know because we know it. Your use of the word "groupthink" is curious to say the least; it seems to have little to do with the coining of Eric Arthur Blair; did you by any chance mean "consensus"? Keeping links to the "highest value" (whatever that might mean; you still have not clarified it) by definition does in fact mean trending towards an extreme; what on Earth did you think it meant? Far from making us look more professional, your approach would give us the appearance of didactic incompetents. All we would need to complete the image would be to limit ourselves to the longest, most obscure terminology and most pretentious sentence structure possible. As it happens, in spite of your implied claim, we have not set any but the vaguest expected education level, so you will look in vain among the WP pillars for justification for it. The linking "system" you urge on us is diametrically remote from any decent research tool, and repeating such claims over and over without definition does nothing to improve it. JonRichfield (talk) 18:47, 6 November 2012 (UTC)

It has been well established long before that, irrespective of linking, we write to the lowest common denominator of an English-based high school level reader, period. There are articles of technical nature where even writing to that can be difficult, hence we suggest "one level down" there (eg a complex math concept normally understood at a gradudate level is written for an undergraduate understanding) but most of our articles are otherwise designed for the knowledge set one gets from high school. If we can include some remedial text without exhausting too much space, certainly we can include that, but in general, that line sets a point where most topics should start their discussion (eg: for rust we start with the assumption of a high school chemistry background, and don't have to go back and explain what atoms, molecules, protons, neutrons, etc. are.) If you can't agree this as a starting point on writing alone, then there's no way to meet that expectation for linking. Once you start with this point, what becomes high-value links are readily apparent, easy to decide and set a line when to link, and avoids endless edit wars over linking low-value links. --MASEM (t) 12:43, 7 November 2012 (UTC)

@Masem, no, not fullstop. You yourself correctly mention concessions both ways, and I agree with both such classes of concessions, in fact I urge them. However, those concessions have little to do with linking. Where both advanced and naive readers abound, the very same (competent) linking can sometimes be valuable in helping people who otherwise would have had trouble reading an article, for example both by speeding the progress of advanced readers past material that they otherwise would have had to skip and search their way past, and also by directing less technically qualified readers speedily and smoothly to parenthetic material that can enable them to assimilate an article without struggling to find supporting text. In a competently written article the links match the content and the logical information structure, not the aesthetic fancies of certain categories of authors and editors, and especially when affronts to such fancies could easily be circumvented. Nothing that you have said yet amounts to a functional claim that a higher frequency of links should affect the value of the articles to a reader of the appropriate level, though it might in some cases improve the value of an article to less specialised or less highly trained readers in the field in question. Going back to explain atoms, hadrons, leptons, bosons and fermions etc in your somewhat strained example, might be assisted by extra links that graduate physicists, philosophers, and chemists would not require (fat chance in some cases!) but quite a lot of other people who had passed high-school chem years before, might indeed need a bit of assistance with say the effect on rusting, of pH or the presence of particular transition element ions, no? And to avoid bothering our curly little heads about the problems of both advanced and naive readers, we are to ... errr... avoid the use of links to improve the quality of our articles??? I hardly dare raise the issue of culture clashes in joking, Masem; you surely are taking advantage of my humour bypass! Next no doubt you will be wanting to ration the number of places where we may use anything as stridently distracting as citations? (Can't you hear them: "What are all those stupid little blue numbers doing there?" Hmmm??? And a lot fewer folks read citations than links, I bet!) JonRichfield (talk) 16:30, 7 November 2012 (UTC)

People have complained about the citation superscripts being annoying too - often this is where you will see the "(Smith, 2000)" inline cites being used instead. Blue links are distracting if you are trying to read for understanding. Blue links are even more distracting if you have no idea of the value of the link to the topic that are you reading - they are like ads that break the flow of a text article. When you keep linking to the highest value links, then now the links are much less a problem because its clear that if they are needed to be clicked, you are going to find highly relevant information for what you are reading presently. And furthermore, being careful in selecting our links makes us more professional and improves our perception to others, instead of looking like TV Tropes where links are haphazard and can cause people to become sidetracked and lose focus. So, yes, careful link selection drastically improves the quality of our articles within the context of what our target goal is: to be an encyclopedia, a research and reference tool. --MASEM (t) 21:31, 7 November 2012 (UTC)

On a point of morbid curiosity, am I now using the correct convention for indentation? Seems pretty cumbersome to me, but if everyone likes it, OK. Now Masem, does it not strike you as just a little obsessional to make a be-all and end-all of blue links and blue citations, rather than service to the reader? And even more so when you cannot offer any basis for when to draw the line in terms of "level"? High-school chem hm? And high-school geography? What on Earth persuaded you that "high-school" represented a defined standard? The well-known fact that everyone does the same subject in high-school, profits from them identically, and retains such profit identically? Now you want to prune citations as well? Why not distracting facts as well? After all, who the blip wants to know about pH and Arachnids' spigots anyway? Chuck the lot out! And still more, just because you don't like blue, you want the links and citations omitted but militate against permitting readers to avoid distraction by hiding the blue? What's your problem? Are you afraid of being distracted by the thought that a word might have been blue if only it had not been masked according to the taste of the reader? I propose leaving preferences to the reader; you insist leaving tastes to you. Please do not tempt me into once more pointing out the merit of your views on professionality in terms of high value links, that you repeatedly are unable to define because they are indefinable; the concept as you present it is ambiguous and arbitrary. You have no basis for deciding when a link "needs to be clicked on"; there is no such general attribute (right?) Some people need to click on some links. Some prefer not to. You have no cogent criteria to go on. High-school standards may suggest the level of register to aim for (very sorta, very kinda), but they have nothing to do with what to link; that is something that a professional bases on the coherence of the information, not the colour or the high-school textbooks. When you do tempt me to point out such things, people are prone to wag fingers at lack of courtesy to you. Meanwhile you swear that you are free of the same sins, while insulting a lot of us with pontifications on the likes of distraction and professionality. Source for the goose, say I, and no contribution to any part of any solution, such as some other persons have proposed. JonRichfield (talk) 13:35, 8 November 2012 (UTC)

As a point, you are getting really close to making personal attacks so you may want to back off the rhetoric.

Any linking system we choose - excluding linking every word or the other extreme of using no links whatsoever - will be an arbitrary decision. The question is what is the best systems that benefits Wikipedia and its intended readership. Yes, there are benefits to some readers with more inclusive linking, but arguments in the past have favored sparser high-value links that are better suited for research purposes and less toward casual reading. Hence why common geographic terms are typically not linked in prose. Remember, no one is pushing on a link-less Wikipedia like you imply. Just a link density that does not distract from comprehending prose while providing highly valuable relevant information every time a link is clicked.

Just because high school (or the equivalent) education point is picked as a writing standard (again, arbitrarily chosen to best suit WP's needs) doesn't mean we're ignoring or refuting to include information taught below that - you only scrap that surface at most schools. But it is important to consider what the previous 10-12 years of education would normally provide you - including the fact that major world geography has been taught years before and re-iterated on since. Just as one has been taught English and should be familiar with a vast array of words through so many years, they, by the time they reach high school, should know every continent, most major countries, and most major cities. We're not talking about knowledge taught for one test and then left forgotten; these places are common terms in normal everyday vernacular, just as common as many nouns are. Which is why linking them without thought as to their relevance makes them low-value links, and makes us (WP) look unprofessional. --MASEM (t) 14:46, 8 November 2012 (UTC)

"the assumption of a high school chemistry background" is more consistent with unlinking "atoms, molecules, protons, neutrons" than with "hadrons, leptons, bosons and fermions", so I don't understand the point of changing the example while arguing against it. Art LaPella (talk) 22:42, 7 November 2012 (UTC)

Oh come on Art! You are trolling me, aren't you? In case you aren't, may I remind you of the lament that "if you don't give an example, then no one understands you, and if you do give an illustrative example, they think you are arguing the example." <siiigh> The point is that it is quite possible to quibble over the justification for linking at any level. Some folks would say: "Are you crazy? What did you link to such obviosities as atoms and ions for? Errr -- Protons? Well, maybeeee... let's see. Others (including a surprising proportion of those who did pass their high-school chem) would say: "Oh good grief, I never did understand about those silly ions and things; I wanted to know about rust for Pete's sake; what is all that blue about?" And on whose side should the editor be? No competent editor links or refuses to link because he thinks everyone knows (or bloody-well should know) what the editor knows or because he thinks nobody knows, or even needs to know what the editor doesn't know. Nor does he link or refuse to link because he thinks that everyone who doesn't care about the fundamentals should be spared the presence of links while more committed readers could damwell go hang and take their pretentious curiosity with them. The reason to link is to maintain coherence while not crowding text with information that could profitably go into another article, while leaving the decisions on what to read, and when, with the reader. It is not to dictate some magically-determined level that everyone must agree is high-value linkage or something like that. Art, if we do not commit to serving the reader, we could save everyone a huge amount of effort, you know? Just whom did you think WP is for? And if you think it is just conceivable that readers' tastes and educational levels might differ, do you really think that it is more important to spare the advanced reader the trauma of reading past blue citations and links, or to support the less informed reader in expanding his education and competence? Now, have I left you still in doubt about why I illustrated the difficulties of arbitrarily assigning levels of link (ugh!) "value" in terms and in depth of profundity that all would agree on? It seemed to me at the time that the remark on deeper levels of conceivable linkage (leptons, bosons, etc) would be more concise than a discussion along the lines of this note. Silly, silly me! JonRichfield (talk) 13:35, 8 November 2012 (UTC)

Rather than get sidetracked any further on who really meant what, suffice it to say I agree "that it is quite possible to quibble over the justification for linking at any level." But I do know that I'm not "trolling"; that word was reckless. Art LaPella (talk) 00:28, 9 November 2012 (UTC)

And just to expand on the idea of bad trends, at one point we used spoiler warnings for works of fiction on WP. The common sense approach was to use them on current works (a few months out), but then people starting using them on older works (up to a year) claiming that it may take people that long to see/read them. And then it started getting out of hand with spoiler warnings being used on Shakespear plays, fairy tales, and even works of non-fiction. Yes, they are their to help selected readers who don't want to be spoiled but it got in the way of serious research articles on fiction works. Now, we have completely stripped spoiler warnings everywhere in WP which is completely an acceptable level. This similarly can (and in some cases has) happen with low-value links. --MASEM (t) 21:57, 4 November 2012 (UTC)

This sounds altogether reasonable. However kindly the intention of spoiler prevention in such a form or sense though, it also sounds altogether unpractical, even impracticable, so we are at one on that aspect. But I think that the general topic of bad trends arising from good intentions is too wide, too commonplace, and too vague to be relevant to the current topic before anyone has managed any cogent demonstration of an operative isomorphism, which I deny has been achieved hitherto. JonRichfield (talk) 18:47, 6 November 2012 (UTC)

@Tony: I would include the 's into the link (Washington's) otherwise it would look to me as the link indeed goes to the top of the main article.

@everyone: Could anyone please summarize what all this wall of text is about? Is it still the usual “people disagree about how often to link to cities”, which has rounded in circles for years without anybody trying to gather any kind of evidence (e.g. surveying readers)? (And also, couldn't everyone agree to one standard of talk page indentation?)

— A. di M.  11:20, 5 November 2012 (UTC)

@A. di M. Not sure what counts as indentation standards. Not sure anything I have seen looks much more helpful than most others, apart from copying the indentation of the chap whose para you are replying to, which is a bit of a pain. Your @format looks pretty handy so far. JonRichfield (talk) 18:47, 6 November 2012 (UTC)

@A. di M. The apostrophe s you refer to sound right to me too. JonRichfield (talk) 18:47, 6 November 2012 (UTC)

@A. di M. Summarising... Yerrr.... welll... Your point is not just good, but excellent. It certainly goes beyond city/country links as such, addressing the whole field of disagreement on either single or multiple links within any article. Not quite sure that I in particular bear the veritable CHARISMA though. My curt manner is prone to give offence where none was intended. Furthermore, I think that anyone attempting a comprehensive precis at this point would be likely to raise a lot of hackles. Still, the need you point out is undeniable. How about something on these lines:

Someone formats a table and posts it here with three columns (or four, counting a leftmost row number). (Just a suggestion; better formats welcome!) Leftmost column headed Link-endians or Blue-haters or something. Middle col Link-startians or Blues-lovers. 3rd col: Proposal for solution or compromise. Then let the guys/gals do their best to concisely (nice word that; wonder what it means...) enter one point, grievance, or rebuttal in a cell, or add a new row with at least one cell filled in. The idea being that either a rebuttal or enlightening remark (say, acknowledging a difficulty) be placed in the opposing cell. I had been going to suggest that the 3rd col contain a proposed solution, but I recommend that a separate, numbered list of actions, one terse paragraph per row, be appended, and that col 3 just contain relevant numbers, each identifying a solution conceived to meet or mitigate the problem in that row. Maybe when we see the list we will be in better case to agree on some action.

No idea how this will fly, nor even whether it really is necessary, just bouncing it off well-meaning skulls. JonRichfield (talk) 14:28, 6 November 2012 (UTC)

I like the table idea. It'd be much easier to read (theoretically). InedibleHulk (talk) 13:34, 8 November 2012 (UTC)

Protected

Fully protected for a week due to the edit war. If you can find a consensus somewhere on this page that can be easily pointed to then do so, I have read through all/most of the above and couldn't see if there was or not, then feel free to ask any admin to unprotect the page. CambridgeBayWeather (talk) 20:17, 12 November 2012 (UTC)

Archiving

I've manually archived everything up to April 2012. Could someone take a look at the Miszabot code at the top of the page and figure out why it's not working? DoctorKubla (talk) 17:00, 29 October 2012 (UTC)

Finally figured it out. Should be okay now. DoctorKubla (talk) 08:55, 27 November 2012 (UTC)

Just noticed this debate due to Boundlessly's statement of intent to resign from the project

After reviewing the above debate, I am appalled at Masem's position and believe JonRichfield and Boundlessly have fully articulated the better position. Masem has clearly not traveled enough or had enough life experience. (I hate to be a snob about this, but I have visited over 20 U.S. states and 11 foreign countries on five continents.) What Masem doesn't seem to understand is that many intelligent people who can use English quite well at a high school level or better are still operating in dramatically different cultural frames of reference, especially in the context of an encyclopedia with a global audience like Wikipedia. --Coolcaesar (talk) 12:28, 15 November 2012 (UTC)

Chronological items

I was reading the "Chronological items" section and wondered why it didn't have blue links to the articles it was discussing, such as February 24 and 1789. Sure enough, on checking the history I found that the links had been removed in this pair of edits with the edit summary "Remove links to Years per Manual of Style/Linking".

I don't think the MOS applies here, because the text is talking about the Wikipedia articles having these names, and not about the dates and years themselves. I have restored the links. -- John of Reading (talk) 21:13, 24 November 2012 (UTC)

Applause for the deed and yet more applause for the reasoning. Some folk are going to hate you. JonRichfield (talk) 19:33, 27 November 2012 (UTC)

Tread softly for you tread on my national pride -- whatever that might be...

In my time I have encountered a sizable number of cultures at various depths of intimacy and intensities of mutual incomprehension. There may be cultural levels at which sense of humour is so rudimentary as to be practically incomprehensible to Joe Miller, but in the circles at which we correspond I reckon the you might as well forget the idea. That there are cultural incompatibilities is unquestionable (I frequently remember events such as when a lot of us in a South African office were recounting our favourite Goon Show snippets and killing ourselves, and I caught sight of the blankly uncomprehending face of an expat Scandinavian) but these incomprehensions are as a rule symmetrical. It is not cultures that are without humour, but individuals. The fact that cultures and their members do not share their humour globally or indiscriminately need not mean that they are respectively impoverished, just that their contexts don't match. The fundamental mechanics of humour are to a good approximation universal in principle, if not necessarily in mood.

Irony? Sure the Americans have it cut and dried, practically composted; they just often use it in different conventions and situations and moods from those of the British, or the Germans (whose irony, like South African, also is drier than that of most Britons). But even in making those generalizations, I am creating a nonsense that I can't spend enough time on to salvage. The British themselves hardly appreciate the British sense of humour because the British sense of humour is legion, not any one thing. What works in one county might earn you blank or black looks in the next. Quite unlike the Americans, Ozzies, Greeks, Chinese, French or Indians of course. And as for the idea that there might be a Briton without a sense of humour -- perish the joke! JonRichfield (talk) 19:33, 27 November 2012 (UTC)

Raking over a few ashes

Tony, Masem et al. might find it entertaining to contemplate an article in which I have just done a bit of editing of some of my earliest work, produced while I had been brought under the impression that almost any word whose concept appeared in an article should be linked, and in fact, linked frequently. Later I developed a bit more perspective that, in a triumph of hope over overwhelming evidence, I am trying to present in a constructive spirit to some of my fellow-editors. The article is Cassytha. The point I wish to make here is firstly that I had linked names such as Africa, Asia, Australia, and the Americas. Secondly, that I have now unlinked them in the light of subsequent experience. Thirdly, that I left Australasia linked, but am in doubt about whether I should.

Why? Because, looking at the text, it was plain that anyone reading the article would not in context be looking for general info in articles on those continents, and if they did link to those sites they probably would not find Cassytha explicitly mentioned, and if they did happen to find it, it would only be after a long, long search. However, Australasia left me scratching my head. Most educated people would have a more or less broad idea of what it means, but some, such as my late aunt Maggy Youngfly, would be likely to say: "Huuuhhh???" and for the likes of her I made it easy to rid their minds of their doubts at the cost of two clicks.

So why mention Australasia in the first place? What is wrong with vanilla Australia? Because Australasia is a geographical region relevant to the genus.

And if in fact the articles in question had indeed mentioned material relevant to the topic, I should unhesitatingly and unapologetically have linked to them, either under their own names, or under the description of the material in question, or both, no matter how familiar the Great Washed might be with their names.

Even more recently I not only edited Calico cat and Tortoiseshell cat and Cat coat genetics, but actually linked them to each other, even though every educated person is familiar with the topics and their mutual relevance, and I'm glad, glad, glad!!! Maybe someone uneducated will read the articles and become educated! I realise that it is not the function of WP to educate, but I cannot help thinking that a little incidental education can't do too much lasting harm. Can it?

OK guys, am I making sense? Comments? JonRichfield (talk) 18:58, 27 November 2012 (UTC)

Australasia is a reasonable geographic term to link because it is not a commonly used name and as you note its specialized for this specific area. Heck, the only thing in that article that I'd contest is the linking of "common name" (which you have linked twice in the same para), the rest of the links show high levels of germaine or , in cases of words like "vector", the specialized meaning to that field. --MASEM (t) 19:25, 27 November 2012 (UTC)

Oooopsie! Thanks Masem; that was a slip. I have delinked the second one. (Shows how non-distracting the links can be!  ;-) ) And I appreciate the complimentary note. JonRichfield (talk) 08:11, 28 November 2012 (UTC)

So why not spell it out in the text? "Australia and New Zealand", so the reader understands without fuss or diversion? Tony (talk) 03:58, 28 November 2012 (UTC)

Tony... I recommend that you check on the Australasia article. Also, whether one spells out a meaning where one word (possibly linked) does the work of two, one could argue the choice of expression as a matter of taste. Where one word does the work of several in an unambiguous fashion in context, there is reasonable justification for preferring the more compact form. (Sez me anyway!) Anyway, since the word is biogeographically useful and most readers will have no need to use the link, it is a service to those that do use it, to introduce them to it in context, without significant penalty. No? JonRichfield (talk) 08:11, 28 November 2012 (UTC)

One of the problems of linking to Australasia is that the article itself is equivocal in defining the term. That's why it's better to nail it in the anchor text. I think the intended meaning is of just the two countries (which is what many people understand it to be). Tony (talk) 10:50, 28 November 2012 (UTC)

Actually not; You might argue that Tasmania is part of Aus, but it is not mainland Australia, and several of the species of interest occur in New Guinea as well. Those are some of the main reasons that it is convenient to refer to it by one word; and convenience is a virtue; ask any politician or lawyer. (But not any administrator of course. Keep the red tape flying...) JonRichfield (talk) 15:31, 28 November 2012 (UTC)

and link to Flora of Australia instead, as 'Australasia' appears too broad a topic. -- Ohconfucius ^{ping / poke} 04:14, 28 November 2012 (UTC)

Here you lose me. The topic is Cassytha; some of the species are variously Australasian. How did Flora of Australia get into this as a topic? Are we at cross purposes? JonRichfield (talk) 15:31, 28 November 2012 (UTC)

Australasian ecozone? — A. di M.  16:40, 1 December 2012 (UTC)

Have a heart Adim! I am a naive little humourless lad who takes things that people say seriously seriously! JonRichfield (talk) 18:30, 1 December 2012 (UTC)

Wow... Sorry Adim, I misread what you said a while ago and was in a rush. I just happened to pass through this way today and realised that I had boobed and you were quite right. I have now changed that link to match. Thanks muchly. JonRichfield (talk) 21:00, 16 December 2012 (UTC)

"Once per table"

This edit added this rule. However, there was no discussion around this specific change, and if my memory serves me, previous discussions on the issue were not conclusive on this point; I think there was considerable scepticism due to a) the aesthetics of patchy linking in a table format and b) the impact on sortable tables, given that you can't necessarily say which is the "first" occurrence. And, searching as I write, here is the most recent discussion. N-HH talk/edits 12:02, 4 January 2013 (UTC)

Yea, that new line needs some clarification. In sortable tables, repeating links (if links are being used) are necessary since the sort can change which comes first. I can agree that in an unsortable table (which includes the infobox) that the once-per-table is good practice, though with sensible exceptions. (The same reasoning applies to inline cite reference links if used, since reordering cites within the article body can easily reorder which cite goes first.) --MASEM (t) 16:24, 4 January 2013 (UTC)

What Masem says sounds very sensible. Tony (talk) 10:34, 5 January 2013 (UTC)

With apologies for putting pigeons among cats, and thereby departing from a sturdy old tradition, in this case I find myself compelled to point out that the foregoing is largely reasonable and I cannot be expected to maintain my disagreement with everything that is said anywhere. Better luck next time! JonRichfield (talk) 17:25, 6 January 2013 (UTC)

"Unless particularly relevant"

Despite the sarky edit summary, I was not "huffing and puffing" about "a point ... unlikely to elicit dissent". I was merely pointing out that people have not been calling for – and have certainly not agreed to – to the removal of this long-standing, substantive qualification and reverted its removal (I quite happily left all the other changes). They haven't, and in fact it has been the subject of voluminous and moderately acrimonious discussion previously. Consensus does matter you know, however convinced anyone might be that their preferred version is much, much better. I'm also going to revert the latest change you've tried, which is not refinement but a second attempt at removal, albeit with a replacement qualification this time. It is not unreasonable to expect someone to propose a change of this sort on talk first, especially once their first attempt to change it without doing that has been challenged.
The suggestion that something needs to be "helpful to the reader's understanding" has a very different meaning – if it has any meaning at all – from the qualification of being "particularly relevant to the topic of the article". The point there is that while, for example, one would not normally link passing mentions of "dog" in most articles, one might well link it in the page about a poodle, which seems a pretty clear, objective and sensible distinction, one would have thought. How one assesses what millions of different readers might, individually, "understand" or what is "helpful" to them in that respect, I am not sure. Nor am I sure that "helping understanding" is the sole aim of linking. Facilitating navigation between "related information" is also important, as the lead to this guidelines states. N-HH talk/edits 07:45, 18 January 2013 (UTC)

My edit summary said, "It's not necessary to huff and puff about consensus every time a point is made which is unlikely to elicit dissent", though perhaps (in addition to spelling consensus correctly) I should have said "It's not necessary to huff and puff about consensus when making a point which is unlikely to elicit dissent" -- that is, you were making the point that it's important to be clear in warning against overlinking, and it's unlikely there will be significant dissent from that concept. I made an edit which I thought improved that warning; you think maybe the old wording was better. Fine. Why not just focus on the wording instead of yelling consensus right away? For all we know everyone might have happily sat back while you and I, through a series of friendly edits with explanatory summaries (explanatory of thoughts about content, that is -- not about WP policy) came up with wording everyone found satisfactory. If not, then we discuss to consensus. Just calm down, that's all I'm asking. And see below. EEng (talk) 14:40, 18 January 2013 (UTC)

Agreed. Too many articles link every word beyond a primary-school reading level as "helpful to understanding". — kwami (talk) 08:04, 18 January 2013 (UTC)

The first edit did though leave the provision with no qualification, such that the guidelines basically said "never link common words", without any following qualification; ie in one way the proposed changes are looking to restrict linking. I've always been in favour of avoiding such links in most cases, but I think we do need that longstanding exception to allow, or even recommend, such links where they clearly relate directly to the main topic at hand. WP entries do more than simply define words, they provide further and detailed information, which will be worth linking to in some contexts even if the basic concept, thing or event is well known per se – assuming, even, we can define "well known" for a disparate millions-strong readership. And of course what "helping understanding" means is even more impossible to pin down and, as noted, can be cited by those in favour of more linking as much as it can be used to back up the more radical delinking. N-HH talk/edits 08:28, 18 January 2013 (UTC)

My apologies for not having been involved in this bunfight, and I don't expect to make any more progress than before against the weary drone of "...from distractions and duplications of links, good Lord deliver us!" The younger members among us won't remember this, but in the distant past we had carefully reasoned, structured, and compiled lists of when to link, when to duplicate, and in each case, what reasoning to apply, which was not necessarily the same in all cases <shock!> or for all users <horror!> Where would we have been by now if we had not been protected by the valiant cadre who erected a zareba of "everyone who disagrees with us is a head-in-butt compulsive linker who wants every article to be blue from title to tilde (yes! I am pointing at you bowb, you with your finger on that square bracket key!)"? Heaven knows where it all would have ended. How feeble in response is the whimper of "...but links are supposed to be useful, whether duplicated or not...<snvl> ...why not just see whether the linked site is relevant and can or should contain material helpful to anyone likely to be reading this <snff>... why not just offer options for people who come out in pemphigus whenever they see blue and see red whenever they see red, to avoid these horrid contingencies... <boohoo>?" Good Grief folks! Rally round the boma; some articles contain some unnecessary links and our foundations are straining under the burden -- if we do not maintain eternal vigilance we might find ourselves in an actual encyclopedia! JonRichfield (talk) 09:06, 18 January 2013 (UTC)

I agree with N-HH on many matters, but here, I have to take the lead from Kwami instead: helpful to readers (above primary school level indeed) is the operative word for optimising the utility of the wikilinking system. Tony (talk) 09:08, 18 January 2013 (UTC)

BTW, I notice that the drone omits the question (among a lot of other valid and relevant points that are best omitted) that in picking on common words as an automatic target of anti-linking fervour, is to ignore links in which the topic is most smoothly and helpfully supported by a link of a word or phrase that looks like something out of a kindergarten, but links to a relevant, non-trivial topic, whereas a link to a forbiddingly technical term might be useless, because though the word matches the title and the sense, the article does not contain any reference to the concept in context. For example, though chimps occur in Kenya, the word does not appear in the Kenya article, so one should think carefully whether to link to Kenya from a chimp article. I leave it to the reader as an exercise in development of a sense of functionality, to decide when it might be worth linking to an article that does not mention the subject of the linking article. (Helpful hint to the baffled: "Unless particularly relevant" won't cut it!) JonRichfield (talk) 09:31, 18 January 2013 (UTC)

"Relevant" combines the ideas of connectedness and "appropriate" (whatever that means, which is the problem). The latter is what we want links to be, but I worry most people think of relevant as meaning merely "connected somehow", and that's what we need to work against. If I'm right about this, then we need something more than relevant, or something else. By substituting "serves the reader's understanding" (or whatever) I was attempting to provide a goal in terms of serving the reader, instead of a statement about the abstract relationship between two topics. That's all. I wasn't "pushing to remove" anything [1].

N-HH raises the very real problem of "Who is the audience?" -- a child or a well-informed adult. But I don't think this goes away by appealing to relevance instead of helpfulness -- either is relative to the audience. Honestly I don't think any brief criterion stated here is going to resolve that problem. I was only trying to provide a better statement of principle to inform the zillions of individual linking discussions, which for the moment are going to have to resolve the who-is-the-audience question for themselves, since so far we aren't doing it here (or anywhere else that I'm aware of). I'm not married to it.

How about somthing like this (though please help with the drafting):

In particular, the following are not usually linked unless particularly helpful and relevant to the reader's understanding. Particularly helpful and relevant means that a link exists from Article on Topic X links to Article on Topic Y only where Y contains (or would be expected to contain, once appropriately developed) information which [and here my creative powers are exhausted, as I trip over this already overextended sentence ... can someone jump in?]

EEng (talk) 14:40, 18 January 2013 (UTC)

Everything seems to have gotten a little back to front here. My general position is not to worry that much about "overlinking" and to be happy to see links to related and relevant pages even if they relate to supposedly "well known" things (nonetheless I do support broad guidance against the unthinking mass linking of common terms when mentioned in passing). Plus, Tony, this is not the "consensus version", it is the version EEng added only a few hours ago, which no one – whatever else they think – has yet explicitly backed up or agreed to. Also, I'm confused by your comment about agreeing with me generally but with Kwami here. In fact of course, we tend to disagree about many linking matters (although not, overall, by as much as is sometimes suggested); and in this section, Kwami, AFAICT, was expressing scepticism about "helpful to readers"-type wording, since it can be taken to mean all things to all people and, specifically, would encourage "overlinking" of the sort I know you deprecate. In respect of some other comments, and the point more generally, I have been pretty clear above and previously about the wording I would prefer and why, which, as it happens, is more or less the long-standing qualification re "relevance". If anyone wants a recommendation from me, that's it. I can't see, for example, that the draft suggestion improves much either in terms of clarity or intention. N-HH talk/edits 16:03, 18 January 2013 (UTC)

• I actually prefer "relevant" for the reason that it removes the subjectivity as to what could be considered 'useful' – we've been around the houses here with no real consensus as to what "useful" actually is, but I have often gotten the impression it's been argued by some to add navigational and gloss links to further clutter up the screen. I actually think it might be good to replace "relevant" with "relevant and useful". -- Ohconfucius ^{ping / poke} 02:55, 19 January 2013 (UTC)

That intention is not unreasonable, but the idea of "particularly relevant" is not a lot better than "particularly true" or "particularly unique". Accordingly it is not a good basis for justifying the use of the term in a guideline. Something more like "...particularly likely to be helpful to some readers..." maybe? Whether it is relevant or not should not be an issue; it it is not decidedly relevant, it should not be linked at all, never mind more than once. It can of course happen that a link would be irrelevant in the article in the first place that the term occurs, but that it is relevant in the third place, in which case that is where it should be linked. It also can happen that it is relevant in one connection and shortly after becomes relevant in a different connection in another place near by, in which case it should be linked in as many places as matter. The fact that we don't have an absolute criterion for measurement of the attribute of relevance hardly matters; it is the sort of thing that is routinely decided on by editors as a matter of good sense, either alone or in discussion, and rightly so. Mindless delinking remains an offence against the WP and its users and editors either way. JonRichfield (talk) 12:55, 19 January 2013 (UTC)

"...particularly likely to be helpful to some readers..." – yecch, no thanks! Even more subjective and even more likely to precipitate edit wars. I also share a dislike for "particularly relevant", "particularly true" or "particularly unique". Why don't we go for "germane". -- Ohconfucius ^{ping / poke} 15:29, 19 January 2013 (UTC)

Ye.e.e.sss... You have a point, though in this connection germane is really an uncomfortably close simile to relevant. How about "apposite" or "apt"? Both of them have the "relevant" meaning, but also have the meanings and usages along the line of "appropriate" or "fitting" or even of "the very ticket", supplying what is needed, which was the main thing I felt to be lacking in "relevant". They correspondingly also have the merit that there is nothing uncomfortable about using either with relative terms like "more" or "most" or "highly". Of course you might consider "appropriate" itself, but that strikes me as too vague. JonRichfield (talk) 16:17, 19 January 2013 (UTC)

I can't see a huge difference between "particularly relevant" and "germane" (and nor do I see a problem with the former phrase, which is not in the same order of redundancy as "particularly true" or "particularly unique", whose problems are more obvious. Relevance is a relative concept, with differing degrees, in a way that truth and uniqueness are strictly not). Anyway, the point, surely, for each of those is that they are relatively objective, clear and substantive, something that "useful", "appropriate" or "fitting" are not really. The latter are impossibly vague, and such clarity as there is will depend on people's criteria and the context. By contrast – to stick with the dog example I've used before – the term dog is obviously relevant/germane/whatever to the page on Dachshunds, and should surely be an exception to any instruction to avoid the linking of common words; it is not relevant/germane to the page on Karl Marx were that to mention that he first formulated his labour theory of value while walking his dog. And that will be the case whether the reader is a 12 year-old with English as a second language who left school last year and has never seen a dog or a 50 year-old university lecturer and part-time dog-breeder. By contrast, if we were to look to "useful", by one set of criteria, the second dog link might be very useful to the former, at quite a basic level. It might even be useful for the latter, if they happened to want to quickly access and read – or edit, for that matter – the WP page on dog (which does much more than simply say "four-legged furry thing, man's best friend"). N-HH talk/edits 16:36, 19 January 2013 (UTC)

I didn't mean to stir up such a hornets' nest. These judgments are subjective no matter what -- no getting away from that: germane, relevant, helpful. Again, I do think that a standard based on the purpose of links (which, inarguably, is to help the reader) is better than a standard about the abstract strength of relationship between two concepts, ignoring the reader or whether is might help him. But the more I think about how to gauge this the more of a headache I get. Further up the page (WP:BTW) is the advice, Ask yourself, "How likely is it that the reader will also want to read that other article? Maybe that's something like what I'm after, but I feel like that's a few steps' progress on a goal that's still far, far away. Foregoing analogy originally expressed as five feet of progress on a five-mile trip but I can't remember the units-conversion syntax and I don't want to get yelled at as anti-metric or something. MOS discussions put everyone on eggshells.

And we're still stuck with Which readers are we talking about? I suspect a lot of linking debates have their origin in different visions of WP -- as a place of discovery (for children, say, at one extreme) versus a crisp reference work for otherwise well-read individuals (at the other extreme). I vaguely recall technical proposals for links with varying color saturation to indicate how germane (relevant, helpful, childish, on-point, discursive, whatever) each is judged to be; or maybe the reader can select a level of link filtering. But I don't relish arguing over whether a particular link should be designated as this or that "level".

On the other hand, suppose that linking worked exactly as it does now (linkes manually designated via [[ ]]) but on top of that, some sort of automatic linking would link everything else on the page right down to dog and so on; these links would only be apparent as the reader hovered the cursor over a given word or phrase (only words/phrases without [[ ]]-type links). Then a child suddenly inspired to find out more about dogs could do so easily, while grownups aren't distracted by everyting being blue. I'm not really proposing this (and -- believe me -- I fully recognize the technical/computational/definitional problems here) but as a thought experiment, imagine how many debates over what to explicitly link would disappear if all the leftovers were linked in this unobtrusive, automatic way. Just musing.

EEng (talk) 17:02, 19 January 2013 (UTC)

Because there's so many different ways WP could be used or written towards, we have to decide on who the target reader is to normalize our MOS and writing style towards - including how and what we link. This, traditionally, has been a fluent English speaker with the equivalent of some high-school education. Thus, we should be considering links in light of this reader, and that's why terms like "germane" or "relevant" are better than "useful" or "helpful", because we can at least have reasonable discussion of what terms would be germane or relevant to someone with that education level. Terms like "useful" ignore that and have a slippery slope towards linking every noun ever because potentially they all could be useful.

Remember: we have a search bar - we can assume our readers are intelligent enough that if they see an unfamiliar term that isn't linked they can copy and paste it into that search bar and find out more. So it is better to play on the sparser side of links, again emphasizing germane/relevant concepts for linking. --MASEM (t) 17:16, 19 January 2013 (UTC)

My personal linking philosophy is consistent with yours (see [2] and [3] which, though of course the work of many, are linked in a way with which I'm perfectly happy). As to audience, I think that's about right, but... is there a formal guideline on the subject? I can only imagine what those debates must have been like... EEng (talk) 17:56, 19 January 2013 (UTC)

I don't know of any specific past discussion a formalization of a target audience though the point has come up before with this seeming to have consensus. The closest advice is probably listed at WP:TECHNICAL, but we don't specifically spell out what the average reader is. (I also consider that given the goals of the Simple English wiki, there's a reason we can work at this level) Personally, I would think we would do well to have some place to state that, with appropriate exceptions, that we treat the reader as having X schooling and Y years of English behind them, as to help simplify discussions about this. Of course, TECHNICAL provides lots of advice that would override that (eg highly technical articles may not be easily discussed in terms of basic high school knowledge), but most cases when we run into overlinking, for example, are topics of broad understanding. --MASEM (t) 19:44, 19 January 2013 (UTC)

Isn't the whole point that we can – and should – bypass and avoid theoretical debate about putative readers and target audience, not least because there is no way of identifying either of those things? And surely the point about relevance is not about refiguring that whole debate and debating what is supposedly relevant to one reader or another, but about what topics and subjects are relevant to each other? N-HH talk/edits 23:27, 19 January 2013 (UTC)

Arbor tree-ish break

Is it worth beginning a discussion on how to complete the following?

In particular, the following are not usually linked unless a reader who is [insert magic language here] would find them particularly relevant to the topic of the article:

It is recognized that this may put the cat among the pigeons, which is why I want to know how people feel about such an effort before even making a stab at it myself. One problem, for starters, is we certainly can't express the reader's level in terms of US "high school", though as a temporary crutch I suppose we could add "or its equivalent in other countries" or some skimpy fig leaf like that. EEng (talk) 20:33, 19 January 2013 (UTC)

We seem to have a perfectly simple and obvious guideline currently – don't link everything, especially things that might be "well known" at a basic definitional level, but equally, when clearly relevant, it's OK. I'm confused why we are even having this convoluted debate about all that. Let's all go and do something more useful.N-HH talk/edits

Um, OK, you guys go on home. I'll turn out the lights and wash the coffee cups. EEng (talk) 12:46, 24 January 2013 (UTC)

Visually more user-friendly links

I was reading a Swiss gov't historical page. All black, no aggressive, obnoxious blue links, just a faint dash underline under certain words here and there to indicate a link. Nirvana!

Has a more discreet, user-friendly linking format, such as the one mentioned, ever been discussed?--Lubiesque (talk) 16:26, 18 January 2013 (UTC)

The likes have been discussed back in the mesolithic; like every attempt to improve the system they all got shot down by guys who knew that they didn't have to demonstrate anything but that any criticism of the existing standards was a plot on the part of evil spirits who want every word in every article to appear in blue. (Read the archive if you doubt me). They also wanted every link to appear only in their own choice of format and shouted down every suggestion for alternatives or enhancements as forcing other standards on unwilling readers who wanted nothing but blue, and blue that never appeared more than once per article at that. (Incoherent and implausible enough for you? You should have seen the actual discussions!)

Themes that were shouted down included optional user-clickable formats of linking. Some alternatives were:

Options that would not work on mobiles in case they afforded PC users an unfair advantage, for example links that pop up when the cursor hovers over them.

Anything like an on-screen toggle button that could suppress display of links that people of varied tastes find distracting or irritating, such as underlining links, which annoys people who wonder why they were underlined, or blinking or coloured links that cause epilepsy in readers who wonder why they are coloured or blinking, or black links that people see as racist, or any other highlighting that someone might find aggressive or obnoxious, or underlining because it might be overlooked because it is too faint on one screen, or because it is too strident on another.

Any form of hiding links to render repeated links invisible; you see, even if you couldn't see them, concealment like a worm i' the bud still would ... errr... feed on their damask or something.

Anything that looks like pandering to the depraved tastes of users who don't like the existing standards; if you don't like blue, you are subversive. After all, we already have options that we could code in to make links that look like normal text; and readers could try clicking on every word in turn in case it happens to be a link, and if it isn't you can search the page to see whether it isn't a link somewhere else, and if it isn't there either, you can type it into the search box in case you happen to think of the right word for the right explanatory article, and if you fail in that you can always look it up in a book on on google. Flexibility is what it is all about, see?

As for me, I am beyond the pale; I have spent lifetimes in dungeons for suggesting that there be a toggle button somewhere that would permit:

Default as it is, or even like the Swiss Nirvana if you insist, who cares?

Hovering cursor as the current options permit

Links remain operative whether visible or not

Clicking on the toggle button either flips between visible or invisible (but still functional) links, or if someone wants to be fancy, offers various other link format options.

An added option that would be nice would be one that permits one to select any text the reader chooses, then by some suitable click, say a right-click, launches a search for an article with a matching title, or, failing that, for articles containing that text, just as we can do now by typing it in. But don't spend your energy on anything like that; the BROTHERHOOD won't like it. JonRichfield (talk) 08:12, 20 January 2013 (UTC)

Your last point ("added option") is something like what I had in mind here [4], but your approach is cleaner. EEng (talk) 14:57, 25 January 2013 (UTC)

Your question is based on the false premise that using "a faint dash underline" rather than blue links is more user-friendly (presumably, you mean "I don't like blue links"). You have provided no evidence to support your premise; and my view is that the change you request would be confusing for a significant number of our readers. Also, you are quite at liberty to apply your own style sheet, to format links according to your personal preferences. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:18, 20 January 2013 (UTC)

<Ahem!>... JonRichfield (talk) 12:29, 20 January 2013 (UTC)

Oh, and of course, I forgot to mention red links; what about red links huh? What about red links? JonRichfield (talk) 12:29, 20 January 2013 (UTC)

This is about the visual appearance of links, not about what to link, so I guess a more appropriate place to discuss that is MediaWiki talk:Vector.css and/or some village pump. — A. di M.  17:23, 20 January 2013 (UTC)

I don't think so. If a visually less aggressive, more discreet way of indicating a link was to be adopted, then the raison d'être of this whole (very antagonistic) discussion about what to link would subside because normal reading would not be disrupted, as it is now, due to the obnoxiousness of those coarse blue links;)--Lubiesque (talk) 19:03, 20 January 2013 (UTC)

(For the record, I would find underlining more disrupting than colour. YMMV. Note also that certain systems display colours darker or lighter than they should – the small squares in this image ought to be the same brightness as the surrounding large squares, but they often aren't; if in the rightmost squares the centre looks brighter than the surrounding to you, links appear brighter to you than they do in average. — A. di M.  10:37, 23 January 2013 (UTC))

A few years ago, I raised (here, I think) the option of completely redesigning the wikilinking system to make the linked items less disruptive to readers—even of linking every single word in every article, accessible not by normal visual distinction but by hovering the mouse and clicking. I don't dismiss the idea out of hand, but liberal linking, even if it were designed to be seamless visually, loses a significant function, which is to communicate the expert judgement of editors/writers in being selective; in other words, losing the feature of highlighting what those familiar with a topic believe readers are likely to want to click on would be a major disadvantage. Tony (talk) 13:51, 23 January 2013 (UTC)

Tony, the expert judgement of expert editors/writers isn't worth beans when it comes to defining and displaying what us dumb mortals need or should need when we read edifying material. That way vanity publishing lies. The opinion of the person who thinks he needs (or does not want) a link in a given position overrides the opinion even of an expert so wise that he is distracted by any visible link anywhere. The best we can do is to put in links that are likely to be useful/helpful to people who might not understand a relevant term, know a relevant fact, realise where related material of interest is dealt with, or any other items or emphases that links are invaluable for. They also might realise where a link that had not been used several pages before might be of use in a different place, or where a link that had appeared under a different title might be of use under a new title in this new spot. All these things are good, and omitting any such link is bad, bad, bad. Overlinking a lot also is bad -- a little. Some people might find them distracting, especially of the take the form of ugly colours or ugly underlining. Tsk tsk tsk! And because they are so evil, people who like the idea of a duplicate link somewhere else in the same article must not be permitted to put them in, even if they can be hidden. After all, one can do without ANY link whatsoever, can't one? Reprogramming or reformatting ones display to one's taste and typing in and searching for occurrences of errr... which words...? These are such fine impositions for users who have nothing better to do with their time when all they wanted was some encyclopaedic information, afunction that has nothing to do with WP, right? JonRichfield (talk) 09:11, 25 January 2013 (UTC)

I very much like the idea of a toggle button to make the links invisible while still operative. Ordinary online discussion forums have features that allow members who dislike all the bells & whistles (member monikers, pics, logos, etc) to turn them off. Wikipedia should offer the same choice regarding links. Obviously a lot of people would appreciate. OTOH, those who insist on keeping all those links would still be happy. Everybody happy!--Lubiesque (talk) 15:53, 23 January 2013 (UTC)

But Lubiesque, in our civilisation we must accommodate people with specialised tastes. Some are never happy unless they are miserable, and they are not all that happy even then. Others are not happy even when they have made everyone miserable, because it is all everyone else's $%^&* fault for not being happy, which they would be if only they were not so stupid. Such people will then want to punish the people they have made unhappy to force them to be happy the way they should be. So, for example, if it doesn't make you happy to have few enough links to please the guys who know how few links everyone should want, then to teach you a lesson, they not only remove the links that silly people put in that don't know that they don't want, but they also forbid you to have any facilities that hide the links that they don't want in case you get happy for what they can tell are the wrong reasons. Simple, right? JonRichfield (talk) 09:11, 25 January 2013 (UTC)

The biggest laugh of the lot is that WP could have fabulous options to suit every taste at the cost of minor design and coding overheads. If half the effort that has gone into this bunfight could have been applied constructively instead of wasting space, we could have all our cake and eat it already, including searching for links for unlinked text. JonRichfield (talk) 09:11, 25 January 2013 (UTC)

External links in prose

The lead sentence of WP:EL implies that external links should sometimes be used in prose, although unusually (quote):

"Wikipedia articles may include links to web pages outside Wikipedia (external links), but they should not normally be used in the body of an article."

So does the lead sentence of template doc {{registration required}} (quote):

"Use {{registration required}} when you find an external link within a paragraph or a reference citation ..."

(Evidently by mistake, use within bulleted list entries of section External links is neither mentioned nor illustrated. Contrast template doc {{subscription required}}.)

Neither of these pages explains or illustrates where external links in prose are commendable, or even acceptable. This manual page seems to pass over the matter. (seems to me as I skim) --P64 (talk) 19:51, 13 February 2013 (UTC)

Place name linking

I'm a bit confused about how place names should be linked in Wikipedia. What I've seen most frequently is City, State, but I've also seen that City, State is common as well. Because I thought that City, State was the most clear and the most common, I've been changing City, State links to that. Another editor pointed out that this is overlinking and that the correct format is City, State. Which format should be used? This page of the Manual of Style does say that "names of major geographic features and locations" should not be linked, but what constitutes major? Thanks! --qwekiop147 → talk 00:54, 19 February 2013 (UTC)

Well, what constitutes "major" is an eternal question that has never been answered and probably is impossible to answer given the huge numbers of people from different parts of the world who look at WP pages, for all sorts of different reasons and from all sorts of different perspectives and knowledge-bases. Equally, don't forget – as people who cite overlink as a justification for stripping links often seem to – that the guideline carries an exemption that allows links to "major" features when "particularly relevant" to the topic of that page. Subjectivity aside, and on the point in question, I think it's hard to argue that each US state is necessarily well known on a global basis or that the state a place is located in is not "relevant" to that place. As ever though, I guess it depends on exact context, and arguably City, State is neater. Not sure that helps much .. (other than that perhaps to demonstrate again that there are rarely "right" or "wrong" answers to these questions). N-HH talk/edits 10:30, 19 February 2013 (UTC)

We do try to use the wikilinking system in a focused way, especially where there are so-called chain links are concerned—this is almost always the case in target links to city articles, from which the state is linked prominently anyway. Where both items appear side by side I find myself wondering why a reader want to click on Illinois first, rather than the link to Chicago—all the while ignoring the option to read the actual anchor article. MOSLINK also recommends against bunching several links together. Tony (talk) 11:33, 19 February 2013 (UTC)

A frequently-used script that automatically reduces overlinking generally unlinks most world cities but leaves less significant places linked. On this view, there is not usually much point in linking New York, Chicago, Los Angeles, London, Paris, Moscow, Tokyo, Sydney, Singapore, Beijing, etc. Where exactly one should draw the line is a matter for debate. I think in general there is too much overlinking in many articles and you end up with a meaningless sea of blue. Furthermore (and as alluded to by Tony above), in the case of these major world cities I think it is quite superfluous to add the state or country at all, never mind linking it, e.g. Chicago Illinois or Paris France or London England. To me these just sound silly. An exception is Washington DC to avoid confusion with Washington State. To those who say what about Paris Texas or London Ontario, my reply is that if you say "Paris" people will surely assume you are talking about the capital of France unless otherwise stated. -- Alarics (talk) 11:53, 19 February 2013 (UTC)

In reply to Tony, as ever it is not up to one individual to try to divine, far less dictate, what other people looking at WP pages should click on to, or even should want to click on to, to quite that extent; some people might well want to move directly to the state article, which is fairly clearly related and relevant, and likely (eg in the case of Illinois) not to be that familiar to many people by any objective standard, hence such a link is fine by the actual wording of the guideline, even if one or two people, yourself included, would rather not have it. You're just one editor and reader among many. In reply to Alarics, such scripts are not officially mandated in any way, and their use for such blanket delinking is a matter of controversy – the small number of people using them should have long ago secured specific consensus for their use (which they have never done), rather than everyone else simply being expected to not only accept their use but also automatically fall in behind the individual preferences they represent when the terms that are included for delinking come up for discussion in individual cases. As noted, I think the anti-bunching argument is probably stronger than substantive overlinking in justifying a single, neater "City, State," link format, but equally I can see the value – and guideline-based justification – in retaining a separate link to the state. I also broadly agree there can be too much linking in any articles, but I do not think these City, State, cases are anything that anyone needs to get too exercised about one way or the other (even me). N-HH talk/edits 13:26, 19 February 2013 (UTC)

Then why didn't they go to the article on Illinois in the first place. People do actually read our anchor articles, don't they? I hope so. If they just go to articles looking for links to other articles, there's not much point, is there. Tony (talk) 14:06, 19 February 2013 (UTC)

Tony, you shock me! You surely couldn't be suggesting that some people might want to read an article without reading the anchor first? Next you will be telling us that some people wind up in an article without having looked up its headwords, but simply because they happened to click on a <shudder!!!> BLUE LINK!!! No wonder some folks say that linking should be forbidden! And as for anyone who doesn't know what an illinois is, he should be sent straight back to the kitchen to find out. After all everyone in the world, except for the majority, knows about every US state offhand and cares a damn, both about them and and their capitals and their culinary connections, just as every self-respecting US citizen could tell you offhand without linking, which of these are states and which of them are capital cities: Zaranj, Oromia, Jujuy, Jilin, Haryana, Formosa, Chuuk and Bamyan. Right? Even the one that is smaller than Illinois, right? And if they aren't self-respecting, they don't deserve to have links to click on, right? Tsk! I feel so guilty, Tony, having chosen the only eight such examples in existence. Maybe I should find you an easier eight? I tell you what we need, Tony: we need someone decisive to tell us which links are appropriate to be linked once in an article because some reader somewhere who has a high-school diploma might not know all about it. And if he hasn't been to high school, or if it #$%^-well should have been obvious because every American could have told you allll about it blindfolded, then de-link it straight away. Along with Paris, and all the other states, provinces and cities that you have heard of. Like Gamkaskloof, Mpumalanga and Grammadoelas. JonRichfield (talk) 15:08, 19 February 2013 (UTC)

It's standard concept of germane linking. If you say "Person X was born in City, State", the germane link is the city since that's the most narrowest geographic feature and thus the most relevant to the discussion about the person (in this case); the link to the state (or providence, or country, or the like) is not as it is much broader. If the reader is really interested in the state (or whatever), they can get there through the City article (where this should be linked), but the state link is not germane in most cases that follow the example above. --MASEM (t) 15:13, 19 February 2013 (UTC)

I'm no fan of using two or more links side by side when one will do. In the case of most US locations, the most natural link to create is [[City, State]]to arrive directly at the article. In the relevant articles, "[[Cordova, Alaska]]" does the job more cleanly and elegantly than the 'chained' ""[[Cordova, Alaska|Cordova]], [[Alaska]]", although ""[[Cordova, Alaska|Cordova]], Alaska" is fine by me too. because it creates less blue although it may leave room for the misguided to add a link to "Alaska". ""Chicago" is a more 'natural' link to make (so you will have, [[Chicago]] Illinois" instead of "[[Chicago, Illinois]]", which is a redirect. Oh, because it is so well known, there are considerably fewer circumstances where 'New York City' or 'New York, New York', need to be linked, IMHO. And if "USA" is part of any of those sequences, I would definitely not link it. -- Ohconfucius ^{ping / poke} 14:31, 19 February 2013 (UTC)

In reply to N-HH, I have a thing caled "Delink common terms" under "Toolbox" when I go into edit mode. Where does it say that I "should have long ago secured specific consensus" for its use? I have been here four years and this is the first I have heard of any such requirement. If we are not supposed to use it, why is it there? Presumably somebody must have approved it or it would not be available. -- Alarics (talk) 15:29, 19 February 2013 (UTC)

Alarics, I'm pleased to hear that you're interested in this type of housecleaning. We need more editors to do this, so please continue. (Do scrutinise the diff that is automatically provided, to ensure that there are no "false-positive" unlinkings; in particular, watch for inline lists of country-names, and in tables, where the normal result can end up patchy ... often it's best to manually relink in those cases.) Linking for the optimum benefit of our readers, just like writing good prose, is a skill; so is removing the excessive linking that characterised en.WP for its early years. As an aside, basic overlinking is popping up continually when editors dump articles from other language WPs—many of which have no proper linking guidelines. We can only weed the garden, using as much skill as we can muster. Cheers. Tony (talk) 07:01, 20 February 2013 (UTC)

@Alarics, I don't mean of course that people need some kind of licence or ticket before they can make changes including removing links – often similar changes on multiple pages – as part of normal editing. I'm talking about the people running scripts through hundreds of pages a day, automatically stripping out links to self-deemed "common terms", without thought for context. I do think those kinds of wide-ranging and relatively unthinking, semi-automated sweeps, which impose a rigid – and minimalist, reductive – standard based on the individual preferences of the editor running the script, need a bit more prior consultation and approval; which they have never had. That just seems to me to be a common-sense courtesy, if nothing else, in a wiki environment. I'm not entirely opposed to the use of tools to remove the large number of repetitive and tangential links that can genuinely clutter pages, but often it seems people using them do not do what Tony suggests above and apply a bit of post-stripping discretion in favour of reinsertion where relevant or useful in the specific context.

As to the specific question asked, it seems from all the above that consensus accepts the tighter "City, State,"-style link, even if it is for different reasons and with differing levels of "whatever"-ness. As to the wider question, although I'm broadly in favour of more focused linking and the manual thinning out of links in many articles, I can only side with Jon Richfield and repeat yet again my slight befuddlement at the vigour with which some obsess over terminating perceived "overlinking", as well as at the claims that they alone know what is "best for our readers" when it comes to linking and at the need to stamp out the threat of people navigating their way round WP articles in a different way from that which we insist they have to do it. N-HH talk/edits 11:09, 20 February 2013 (UTC)

I have never run scripts through hundreds of pages a day and I always check the actual output before pressing send. I don't know why you should have assumed otherwise. You can check my edit history. Obsessive unthinking delinking is bad but so is obsessive unthinking linking. So many wikilinks when you stop and think about them not only serve no useful purpose, they actually devalue and dilute the whole purpose of linking. -- Alarics (talk) 13:31, 20 February 2013 (UTC)

On this point I cannot but agree. Sometimes one finds a need for a link apart from the obvious ones; if you are writing about a point that lots of people have never heard about, or if even experts in the field might want to inspect a discussion on the point, and the material is not a discussion that has any place in the article from which one is linking, then by all means link. There are many reasons to link apart from the likes of "I bet YOU have never heard of Paris, France, you yob you!" Sometimes there isn't an article that is suitable, but the author simply links to a useful-sounding title without checking whether it covers the topic, or even mentions it. If you are writing about plaster of Paris and the Paris article doesn't mention plaster of Paris, it is no good linking to Paris unless you also update the Paris article to include the relevant material. If there is no article on Plaster of paris in WP (there is actually, but never mind that here) then either do without a link, or create a red link and if possible, write a stub if not an article. If the article does exist, but does not contain the material that the link requires, then either omit the link altogether or supplement the article. There is a lot more to linking than finding a plausible name; we all must have followed a link and found that the target article contained no relevant material whatsoever. Conversely, if an article exists but under a non-suitable name, then link under a suitable name; in fact, if desirable, link to the same article several times in rapid succession under different names if relevant. And as often as the link will be helpful and constructive, include it no matter how many editors raise a whimper or even a howl. Function and lack of function are what matter. Anyone who can't handle that... JonRichfield (talk) 12:01, 28 February 2013 (UTC)

Btw Alarics (as a belated reply to your comment above after JR's comment appeared as a recent change on my watchlist) I didn't say, imply or even assume you had ever run scripts through hundreds of pages a day; I merely made a general reference to those few people that do – who I maintain do need to back off a bit in some respects. As for the "Delink common terms" option you mentioned in your toolbox in your previous comment, I have no idea why you have that. I certainly don't, and I'm not aware it's a standard fitting. N-HH talk/edits 20:12, 28 February 2013 (UTC)

And as another BTW, nothing in my remark had anything to do with anyone implying anything about anyone else's practices or not; I was referring strictly to the remark that "...many wikilinks when you stop and think about them not only serve no useful purpose..." and pointing out that there are other linking sins than just overlinking, and they are a lot more sinful than overlinking, which is in fact venial, as I have remarked before in various contexts. JonRichfield (talk) 10:29, 1 March 2013 (UTC)

Image Link Policy

Where can I find style guidelines covering when to use an image as a link? e.g.[[Image:foo.jpg|link=Foo Article]] I see images used as links to good effect on plenty of portal pages, but rarely elsewhere (maps, flags, a few other cases). Where can I find info on proper use of images as links? (I've read a number of MoS pages without success so far, sorry if I have overlooked the obvious here). Thanks! -Dschwar12 (talk) 16:26, 27 February 2013 (UTC)

I don't believe we have explicit policy but it seems consistent that in main space, images that aren't image maps should never have links - that is, clicking on the image should always take you to the file: page for the image. If there is a term to be linked associated with the image, it should be in its caption. But this is not explicitly given, nor necessarily consensus, just common practice. --MASEM (t) 15:39, 1 March 2013 (UTC)

Just how 'common' is "football"?

Just how 'common' is "football", or should we consider it to be on the par with "United States" when it comes to linking?

Most often, I see article text that simply says 'football', and it is piped linked to 'association football'. That suggests to me that it is frequently used to gloss or disambiguate the term using in its abbreviated form. Yet the sport is undisputedly "the biggest game in town". It's universally known; the most popular sport in the entire world, and is played at all levels in all countries. Should this term still enjoy systematic linking, and in that manner, or can we rely on other links or clues, or ever writing out the term representing the specific football code? -- Ohconfucius ^{ping / poke} 03:06, 2 May 2013 (UTC)

Because "football" by itself and without context can have two different meanings (association and American football), linking depends on how obvious the proper form of the term should be. For example, if we're using the term on an article on an American person where it is mentioned they may have played football in their youth, its reasonably obvious this is American football; reverse is true if the person was British. But if there's no obvious way the term can be taken, linking should be used to provide clarity if it is really needed. --MASEM (t) 03:26, 2 May 2013 (UTC)

Concur with Masem. A quick review of the article "Football," without disambiguation, reveals a multiplicity of commonly played forms of "football" around the world beyond association football and what we call "American football" here on Wikipedia. One could assume that a British or European youth plays association football, and an American or Canadian plays gridiron/American football, but that would ignore all of the other football variants such as rugby, Gaelic, Australian rules, etc. Even as common as association football is, I don't think we can automatically assume any non-American/non-Canadian reference to "football" refers to association football. Dirtlawyer1 (talk) 04:10, 2 May 2013 (UTC)

Let me clarify: the term 'football' is often used piped to [[association football|football]]. Let's first get rid of the ambiguity, then ask just how common 'association football' is. Possible unlinking situations I have in mind could be [[association football|football]], or [[association football]] to 'association football'; or [[soccer]] or [[association football|soccer]] to 'soccer'. Could these examples be envisaged in terms of 'common terms'? -- Ohconfucius ^{ping / poke} 06:26, 2 May 2013 (UTC)

So is it clunky to spell it out without pipe? Tony (talk) 06:51, 2 May 2013 (UTC)ß

It is probably better in the long run to spell out, the first instance of the use of the word "football" in the body, that if it is association or American, and then judging the link's value from there, and then all subsequent uses of "football" can be assumed to be the same, unless of course both types are present; that it is to link to establish the meaning. At most, we're talking one extra word in the article's entire prose. This is even in the case where the specific meaning should be obvious, as it doesn't hurt there. --MASEM (t) 14:03, 2 May 2013 (UTC)

The "Piped links" section of this guideline says: "However, make sure that it is still clear what the link refers to without having to follow the link." If it isn't obvious which type of football is being referred to, spell it out in the article. Wikilinks shouldn't be created solely to provide clarity. DoctorKubla (talk) 08:43, 2 May 2013 (UTC)

It would surely look a little strange and be fairly pointless to have to write out "association football", or use the word "soccer", for every mention of the game in pages on teams, players, notable games etc, eg "John Smith is an English association football player". Yes it might be clearer for a US reader, but it would look fairly odd to many South American, European, Asian or African readers. Sometimes you just need to go with the context and, as noted, linking can provide any necessary further clarity. Also, on the linking point, let's not go down the road of arguing that, as a common term, links to it can be expunged from any and every article. Pages that directly relate to the game can justify retaining a link, even if it is perhaps the most well known game in the world. Quite apart from the point above, the details and history of the sport are not necessarily known universally (especially in, say, the US) and, as we all know, the guidelines and general practice accept and even recommend keeping such links when the linked topic is directly relevant to the topic under discussion. N-HH talk/edits 08:57, 2 May 2013 (UTC)

It's a problem, and I don't quite know the answer. Certainly I spend a good deal of time fiddling with deceptive piping to make it clearer to readers; but I agree that spelling it out isn't going to be the right way in every case. Any ideas about where and where not? Tony (talk) 09:03, 2 May 2013 (UTC)

I did say "if it isn't obvious". In an article about an English football player, I'd assume they play association football, so there'd be no need to specify. And I wasn't arguing that football articles should never be linked to; the appropriateness of such links depends on context and should be decided on a case by case basis, like any other links. I'm just saying that the purpose of the link, as Masem seemed to suggest, shouldn't be solely to provide clarity which isn't provided in the article. DoctorKubla (talk) 09:10, 2 May 2013 (UTC)

• I'm not talking about systematically unlinking piped or direct links to 'association football' either. I wanted to explore the interpretation, and think about what to do in circumstances such as 'Jose Lopez is a [[Spain|Spanish]] [[Association football|football]] [[Defender (association football)|defender]]'. I believe it's unsatisfactory to allow such chained links, but I'm wondering how and to what extent these can or should be rationalised. -- Ohconfucius ^{ping / poke} 09:48, 2 May 2013 (UTC)

• I agree. I'm particularly suspicious about relying on readers to click on a link for basic meanings in the flow of the text. Tony (talk) 09:58, 2 May 2013 (UTC)

I consider "football" the opposite of "United States", at least for readers of the English WP in the U.S. Whenever I see a statement like "John Doe is a football player", I always wonder "Hmm, is that American football, or is this a soccer player?" I believe it *always* helps U.S. Americans to have a link. In the above example, I think many Americans will legitimately wonder whether Jose Lopez is a Spanish person who plays soccer or a Spanish person who plays American football (perhaps in the U.S.). A simple link, as written above, will clarify the statement unambiguously for those who are curious. Yes, the links look ungainly in the code, and chained links are non-ideal, but I think the resulting WP text is desirable. Jonesey95 (talk) 13:05, 2 May 2013 (UTC)

I'm sorry that I'm arriving about a week late to the dance.

I am an overlinking cop and I hesitate when seeing [[Association football|footballer]] used in the lede sentence of player's article. It feels wrong to leave it when I'm unlinking [[England|English]], [[United States|American]] or [[German people|German]] in those same articles, however I do understand that it does give context to the term. However, so does the infobox. The various forms of football all have distinct infoboxes and perhaps that is sufficient to all those who are wondering what variant is being discussed (except perhaps those who are using screen readers).

The term is common, but because the term has variants based on the type of English being used, I think it's best to leave a piped link in place for this term.

And for the record, Canadian kids generally don't play American football, they play the Canadian variant. Walter Görlitz (talk) 21:14, 22 May 2013 (UTC)

Fobarrian people

I have been reverting links to X people, such as here to Somali people, as overlinking. Is that the case or should we be making this allowance? While not the same as linking to the major geographic area of Somalia, it's a way that some editors create the same effect, quite often in the first sentence. Also, if the reason to do so is to inform readers of other X people, then it may not be an appropriate link since the subject may not even be listed at the linked article. Suggestions and advice? Walter Görlitz (talk) 20:33, 22 May 2013 (UTC)

"Somali people" is only linked to once in the article, in that passage where the subject's partial ancestry is indicated. WP:OVERLINK is also clear that it applies to "everyday words"/"geographical areas"/"units of measurement"/"dates". The term in question is none of the foregoing. Middayexpress (talk) 20:40, 22 May 2013 (UTC)

Only one link implies that you're confusing REPEATLINK not OVERLINK. This is a link to a term that people would easily understand in context and so I believe that OVERLINK applies. Walter Görlitz (talk) 21:00, 22 May 2013 (UTC)

I fail to see in what possible way any of the provisions in overlink, as currently formulated, apply to the description "Somali" or even to links to "Somalia" itself. The latter, let alone the former, is not a "major" geographic feature or location – by any standard, surely, however one exactly chooses to define that somewhat subjective term (with due apologies to Somalia and the Somali people) – plus is of course often going to be perfectly "relevant to the topic of the article". Even if you personally happen to "believe" that the overlink guideline might apply, can't you just leave one or two links be rather than imposing your personal, restrictive preferences on the thousands of readers who might possibly find such links useful? I'm sure you and others can cope with one stray blue word if you really have to, and that even those of us who might never wish to click on the link can avoid being forced somehow to do so against our wishes. WP:OVERLINK says nothing about giving random individual editors the right to delete links that they, arbitrarily, happen not to like or which some readers, funnily enough, might never use. N-HH talk/edits 22:38, 22 May 2013 (UTC)

The phrase here is "everyday words understood by most readers in context" and it's not just Somali people, but English people, Italian-Americans and other similar links. I'm sorry that I didn't spell that out. As for arbitrary, that's what I'm trying to determine, why should these be linked arbitrarily at all? Walter Görlitz (talk) 23:00, 22 May 2013 (UTC)

The original query was framed, in some detail, with reference to the "major geographic .." point in overlink. As for the prior bullet point you're quoting now, surely that refers to common nouns and verbs used in passing, things like "country", "table", "house", "was born", "drove" etc? We're talking about proper nouns here and links to more substantive encyclopedic topics from related pages and where additional information on the second topic might be useful to some at least. Sure some – although not all, by any means – people might know what "Somali people" are, but most won't know much in the way of detail about the topic.

My personal view, leaving aside the fact that a lot of these "people" articles are pretty rubbish, is that you can make a case for limiting such links, especially in the first line of thousands of bios (as opposed to, say, articles about ethnography or demography, specific countries, or other "people" groups), under the [lack of] relevance point or, arguably, when it comes to the "bigger" nations particularly, by association with the geographic point. But I'm really not sure how strong that case is and I am not sure how and where you would draw the line and define "bigger" nations.

More generally, as ever, I find myself slightly bemused by an apparent rush to turn the potential-link articles into orphans and to make life more difficult for people trying to use a basic function of the site – linking – to get around to different but related topics. The benefits of having links for those who might want to use them is obvious; the benefits of not having them, or removing them to such a radical extent, even for those who won't want to use them, is harder to see. N-HH talk/edits 08:49, 23 May 2013 (UTC)

• Linking "Somali people" looks OK to me—unlike linking "French people" or "Canadian people" or "Chinese people". Tony (talk) 09:09, 23 May 2013 (UTC)

But what underlying principle, or part of the guideline, justifies making such a distinction and how does one work out where the threshold for linkability falls? As I say, I'm not actually unsympathetic to the idea that endlessly scattering such links, especially in individual bios, is often rather pointless, not least because it will make too much of an individual's ethnic or national identity, but we seem to be relying a bit too much on the inevitably random preferences of individual editors. N-HH talk/edits 11:05, 23 May 2013 (UTC)

It's a good point. Finding that boundary sometimes leads us to a grey area where it could go either way. That doesn't discredit the notion of the binary decision (link or don't link); it just means we have to be thoughtful and careful in weighing up context, utility, focus, etc. Isn't writing style more generally like this? I'm sometimes struck by how we manage to sort out article text at all among so many different editors; but it usually happens, and often quite well, despite all of those grey areas. Somali people is not in the grey area for me, and neither are French or Russian or Canadian people—but it does slightly depend on context. Perhaps it's what makes wikilinking a potentially skilled art, layered on top of the normal requirements of writing good prose. Tony (talk) 13:10, 23 May 2013 (UTC)

I tend to remove all such links when it's a national group, but not when it's not. So [[American people]] down to [[Tongan people]] would be unlinked, while [[Guaraní people]] would not be. Is that inconsistent or acceptable? Walter Görlitz (talk) 14:25, 23 May 2013 (UTC)

Somali people in this instance is an ethnicity, not a national group. Middayexpress (talk) 14:36, 23 May 2013 (UTC)

Subsection

I think we need info on when to link to a main article with the

Main article: ....

format at the top of subsections. Pass a Method talk 13:24, 23 May 2013 (UTC)

Codifying guideline on wikilinks in reference citations

(t - c) 03:36, 28 May 2013 (UTC)

 – Actually, it seems that this was resolved by adding footnotes to the WP:OVERLINK section, per Wikipedia_talk:Manual_of_Style/Linking/Archive_13#Repeat_linking_in_reference_sections. II

There's been a fair amount of discussion on this (see Wikipedia_talk:Manual_of_Style/Linking/Archive_13#Repeat_linking_in_reference_sections in July 2012 and Wikipedia_talk:Manual_of_Style/Linking/Archive_11#Does_overlinking_apply_to_the_references_section.3F in 2011) but I had to search the archive to find it. Is anyone up for codifying the consensus which was established in these sections, which appears to be that the rule isn't strict? I've ran into someone who seems to be particularly fond of guidelines so it would be nice to point it out. II | (t - c) 18:25, 22 May 2013 (UTC)

terms inside quotes

A discussion on linking terms appearing within quotations was just held at Wikipedia_talk:Manual_of_Style#.22Wikipedia_has_an_article_on....22. It is resolved. It is referenced here because, as it developed, it was probably more relevant to this guideline subpage. Nick Levinson (talk) 16:04, 28 May 2013 (UTC)

REPEATLINK in tables

There has been some changes to WP:OVERLINK.

As of December 25, 2012, OVERLINK included the following text:

Generally, a link should appear only once in an article, but if helpful for readers, links may be repeated in infoboxes, tables, image captions, footnotes, and at the first occurrence after the lead. [5]

On December 31, this was then changed (with the edit summary of "clarify") to:

Generally, a link should appear only once in an article, but links may be repeated in infoboxes, tables, image captions, footnotes, and at the first occurrence after the lead. They should still not be linked at every occurrence, but, for example, once per infobox or table. (emphasis added by me).

The wording was subsquently altered again on January 17, 2013 to:

A term should be linked, generally, at most once in an article's lead, perhaps once again in the main article body, and perhaps once at first occurrence in each infobox, table, caption, and footnote. Even within these general limits, the choice of whether or not to repeat a link should consider whether the added value of linking a particular occurence outweighs the consequent dilution of the value of other links.

• Can someone please tell me if there was a discussion about these changes? The first change didn't just "clarify" the guideline, it fundamentally altered it. It went from effectively allowing repeat links in tables to prohibiting them, which has put hundreds of sports tables at odds with the guideline. Many tables/footnotes repeat links since they are not read like prose; indeed, in the case of a footnote the "first" footnote is whichever one you click on. In an indexical table you may read off a particular row. The fundamental essence of the guideline has been altered under the summary of "clarify", and I cannot find any discussion of this alteration. If there hasn't been one then I am going to restore the previous wording; if there has then I think we need to re-think the alteration, because the alteration was mislabelled anyway. Betty Logan (talk) 07:31, 7 May 2013 (UTC)

There have definitely been discussions about this, although I am not sure whether they attached directly to and supported the changes highlighted above. I would also say that the original version noted above seems to be the clearest and most concise. Yes, it has an element of subjectivity and lack of clarity, eg: "if helpful for readers" (In what way? As defined by who? etc) but despite all the words and qualifications added to it by subsequent edits I'm not sure we have any more certainty or clarity now. The final/current wording is just full of waffle tbh, eg "generally" .. "perhaps" .. "should consider whether" .. "dilution of the value of other links". What does the last phrase, which often crops up in linking discussions, even mean?

My preference would be to stick to something closer to the older, more concise wording. We don't need absolute conformity in linking practice – even if such a thing were possible – and people can apply a bit of common sense in each case, without linking everything or, alternatively, rigidly stripping out every link but one. In tables especially, surely we can be a bit looser about "over"linking (which I seem to recall was the broad consensus when this did come up). It is not the end of the world if a page has a few, or even a lot, more repeated links than one or two editors would personally like; that seems far less of a problem than hobbling the navigability and functionality of the site for users who are only looking at one part of a page or at a specific part of a big table. N-HH talk/edits 10:00, 7 May 2013 (UTC)

There seems to have been a bit of WP:CREEP. -- Ohconfucius ^{ping / poke} 10:08, 7 May 2013 (UTC)

I personally find repeat linking in image captions disruptive and unnecessary. What kind of cartoon-comic reader just looks at the pics and doesn't read the adjacent text? Are we presenting a serious knowledge base pitched at 11-year-old boys? Tony (talk) 13:33, 7 May 2013 (UTC)

In what way is it actually disruptive? That's very different from saying that some people don't like or want such links. They don't have to click on them if they don't want to. And no, not everyone is an 11 year-old cartoon-comic reader who only looks at pictures, but such people are allowed to use WP, and use it and read it as they see fit. Who are we to tell them they can't? N-HH talk/edits 13:41, 7 May 2013 (UTC)

And now I see that the page has been badly messed around with since January. It needs to go back to the consensus version. Tony (talk) 13:42, 7 May 2013 (UTC)

On reflection, there are of course two related but separate issues here: a "repeat" of a link, in a table, possibly only once, when that term is already linked in prior text and multiple "repeats" of a link within a table, regardless of what is or is not there elsewhere in the article. The original formulation is arguably confusing in terms of mixing the two up, and presumably that's why the first edit noted above said it was to "clarify", which I guess is what it did, even if it also made the rule more rigid (sometimes though open drafting is genuinely about flexibility rather than being confusing). As suggested, I'm sure in previous discussions no one really objected to allowing the first type of repeat, ie between text and table/caption etc, and there was even a surprising amount of agreement on not being overly fussed about multiple repeat links within tables, due to the fact that they weren't running text – so the distraction/sea of blue arguments fell to one side and, even, consistent blueness was felt to be aesthetically better for tables – and that "first occurrence" can be hard to define in sortable tables. Btw, here is the most recent discussion I could find, which in turn links back to a previous discussion (it also suggests that there wasn't any agreement, before or after, for the first change highlighted by the OP). If you search page text for "tables" in Archives 12, 13, 14 etc, that should bring up some additional recent discussions. N-HH talk/edits 13:43, 7 May 2013 (UTC)

As best as I know the latest discussions didn't clarify one way or the other, but certainly did not rule out multiple linking in tables and footnotes, particularly in sortable tables. These are not "read" in the same way prose is, where links can be distracting, so there's no issue with multiple linking if the editor wants to do so. But they aren't require to , either. The changes made seem against this consensus of editor preference. --MASEM (t) 13:46, 7 May 2013 (UTC)

I have no objection to avoiding the patchy effect in sortable tables (maybe even in non-sortable)—although if someone can provide an example of where this is inappropriate, we should look at it. I added pre- and post-nominals, wondering why they were removed, and removed what seems to add nothing but clutter: "(expressions, phrases, dialect, etc.)". Dialect? Tony (talk) 13:54, 7 May 2013 (UTC)

I think a case where prior to the table, in prose, you define a lot of terms that are wikilinked there as they are non-obvious, for a column that has a limited number of possible values; thus subsequently in the table, it doesn't make sense to relink each use since the explanation is right there above the table. But if data is presented without any explanation, wikilinking each row may make sense. --MASEM (t) 16:03, 7 May 2013 (UTC)

The whole argument against overlinking is based on not reproducing a link a reader will almost certainly have come across. In prose this is easy to gauge, but it does not apply to footnotes and sortable tables; even non-sortable tables may be used in a non-linear fasion i.e. if you want to look up the 2011 Best Actress Oscar Winner it makes sense to link Meryl Streep even though she is linked in the other years. It's just inconvenient to make readers go searching for a linked version of her name. I am largely indifferent to linking in captions and prefer not to do it myself, but footnotes and tables are a different matter. Anyway I am going to restore the original wording because this is clearly a controversial matter for some editors that has not been adequately discussed, and the edits have attached restrictions that go way beyond clarifying the original guideline. In short WP:OVERLINK should only apply to prose I feel, with exceptions for footnotes, and tables really depending on the context i.e. a table with half a dozen rows can probably observe the rule, while a table that requires scrolling and is likely to be randomly accessed probably should be exempt. Betty Logan (talk) 06:06, 8 May 2013 (UTC)

What you've restored is better written. I have to think very hard to work out the practical difference in meaning between the versions, though. Tony (talk) 09:42, 8 May 2013 (UTC)

I was about to start a discussion on this topic (if there wasn't already one). There is definitely some confusion on this matter. Some editors think that it is saying that repeatlinks/overlinks doesn't apply to tables, while others (like me) think that it's the same as for infoboxes, where (as we all know) only the first occurrence is linked. And so, it's almost inevitable that edit warring ensues. Personally, I think it's better not to have any links in tables (if they're found prior in prose) than to have the same ones repeated over and over. --Musdan77 (talk) 22:58, 6 June 2013 (UTC)

I have to agree, as long as the useful and focused links are not too distant from the table. But if sortable tables are linked, all or nothing for a column, please. Infoboxes: no repeats, or many visitors will think that blue is the default colour for infobox text, and won't realise that links are present. Tony (talk) 00:39, 7 June 2013 (UTC)

I agree that sortable tables is where links can be repeated but it needs to be consistent throughout the table (though if this creates a heavy weight of red links, those can be dropped). In non-sortable tables and lists, the repeating of links should be avoided. --MASEM (t) 01:18, 7 June 2013 (UTC)

Direct linking

I've been using WP:AWB to fix, among other things, links like [[fixed-wing aircraft|aeroplane]] to [[aeroplane]] and [[fixed-wing aircraft|airplane]] to [[airplane]] (see this diff as one of many examples). Am I right in assuming that if we're talking about an airplane/aeroplane, which is a type (but only a type, not a synonym) of fixed-wing aircraft, the wikilink needs to go directly to airplane or a title that redirects to it? If I'm writing an article about some random topic and I mention a duck, I don't link to it as [[bird|duck]] but rather as duck. Am I right to do this? Red Slash 20:10, 29 May 2013 (UTC)

Yes, now we have the airplane equivalent of a Duck article we now need to update all the old fixed-wing yadda Bird links that specifically concern ducks. Go to it, pal. — Cheers, Steelpillow (Talk) 11:38, 30 May 2013 (UTC)

asking for a better way to link quoted terms

Is there a clearer way to link words within quotations than (1) adding an {{Efn}} template for each term, (2) annotating each Efn template with the relevant subject of the link's destination, and (3) creating a separate Notes section with a {{Notelist}} template to immediately precede a See Also section? Examples are wife selling, with many links in the Notes section, and child-selling, with only one link in the Notes section. What I don't like:

• the vagueness about why many of the links are in the article, because the Notes section doesn't explain that they are due to use in quotes, which could tempt editors to delete them when they belong,
• even less clarity about why a link is present when the quotation is inside a regular footnote,
• the creation of a separate section when an article may already have a References and Notes section, because bibliographic and discursive notes may already have been combined and separating them may create other editorial problems,
• the placement of the Notes section before the See Also section, when the MOS guideline seems to prefer that they be be the other way around, and
• that other groupings of ref elements are discouraged, which may be relevant if Efn templates are already in use for another purpose.

I implemented the above method pursuant to a discussion at MOS talk (and see MOS Linking talk). I had previously asked how to link quoted terms, when the methods were less constraining than they are now.

Thanks for ideas. Nick Levinson (talk) 20:43, 15 June 2013 (UTC)

Linking in tables: ambiguity in the MOS

In the section What generally should not be linked (also WP:OVERLINK) there is the advice Generally, a link should appear only once in an article, but if helpful for readers, links may be repeated in infoboxes, tables, image captions, footnotes, and at the first occurrence after the lead.

Editors working on tables have understood this sentence in two different ways: (1) that links may be repeated in different parts of the article so readers don't have to search a long text to find the link, and (2) that every instance can be linked. An example of the first interpretation is List of operettas by Offenbach. An example of the latter is List of compositions by Anton Bruckner, which actually repeats red links as well as blue ones. Which interpretation is correct? How can we make the MOS unambiguous? Thank you. --Kleinzach 07:03, 24 June 2013 (UTC)

From the discussions I've read, I think the general feeling is that if a word or phrase in a table is linked, every instance of it should be linked – because a) most tables are sortable, so you don't know what order it's being read in, and b) most people skim tables, rather than reading them from top to bottom – but if the links are provided elsewhere in the article, they don't need to be present in the table at all; for example, List of operettas by Offenbach has links to each genre in the section immediately above the table, so they don't need to be linked to in the table as well. If I'm right that there's general agreement about this, I think a MOS sub-section about linking in tables would be tremendously helpful. DoctorKubla (talk) 07:38, 24 June 2013 (UTC)

Thank you. I'd certainly like to get this clarified. Personally I don't like banks of redlines (or blue links) especially of common words like piano and tenor which must be familiar to almost every reader of what are quite technical pages.--Kleinzach 08:00, 24 June 2013 (UTC)

I agree with DoctorKubla's and Kleinzach's sentiments. Non-trivial items should be linked every time because of a table's sortability; it's of course sometimes a matter of dispute what is non-trivial, but some things must be left to editorial judgement or discussions on the relevant talk pages. -- Michael Bednarek (talk) 11:56, 26 June 2013 (UTC)

Can either of you suggest a revision of the present text to make it clearer? --Kleinzach 15:39, 26 June 2013 (UTC)

I don't think the current wording is ambiguous. It supports repeated links in tables "if helpful for readers"; anything more specific will only create problems. I can't see how the wording can be construed to support "(2) that every instance be linked" – not outside "infoboxes, tables, image captions, footnotes, and at the first occurrence after the lead". In short, I don't see the need for a more specific wording. -- Michael Bednarek (talk) 06:23, 27 June 2013 (UTC)

Do you think the linking of List of compositions by Anton Bruckner is OK then? Specifically the red links to Weltliches Chorwerk? --Kleinzach 08:28, 27 June 2013 (UTC)

No; I think it's obvious that those red links are not helpful. I also doubt that "Weltliches Chorwerk" is a genre; to me, it's a description where the genre is undetermined. That applies in my opinion to most of the genre terms in that list, but I have no inclination to dispute the issue. -- Michael Bednarek (talk) 05:27, 28 June 2013 (UTC)

Yes, Weltliches Chorwerk was really a holding category. It was discussed here . I've started a discussion on genres at CM but I think it may not go anywhere. Kleinzach 06:44, 28 June 2013 (UTC)

LInking to a section inside the same article

Is it correct form to link to a section in the same article as the reference? For example in the Snoop Dogg article, there is a line in the Snoop Dogg#1992–97: Death Row, Doggystyle and Tha Doggfather section which mentions "A short film about Snoop Dogg's murder trial". A full discussion of the murder trial isn't covered until the Snoop Dogg#Legal incidents section half way through the long article.

Would it be acceptable to link from the first mention to the detailed section? Qexter (talk) 17:18, 17 July 2013 (UTC)

Innovative use of piping on Symbol (chemical element)

In many (most?) browsers, hovering the cursor over a piped link will cause the linked name before the pipe, rather than the display name after the pipe, to appear. Symbol (chemical element) exploits this behavior to include the words from which chemical names are derived, and their English translations, in a table.

A note above the table reads, "NOTE: Placing one's cursor over the entry in the etymology column will display a modern English translation for words of Greek, Latin, or other origin." In the table there are links such as [[inert|'''''ar'''gon'']] which display as argon and for many readers pop-up inert. (For other readers, I guess it constitutes an Easter egg link.)

The problem I had with this is that a number of the links were to DAB pages; I changed several to piped links such as [[Bearer (carrier)|phoros]] or Wiktionary links such as [[wikt:escape|'''''la'''nthano'']]. I wonder, though, what you manual-of-style types make of this. Cnilep (talk) 07:28, 29 June 2013 (UTC)

Definitely an improper use of piping. I don't know if there's a way to generate more conventional tooltips using wikicode, but there's plenty of room in the etymology column – why not simply put the modern English translation in brackets beside each entry? DoctorKubla (talk) 08:07, 29 June 2013 (UTC)

Phrases (and the underlying techniques) like "Placing one's cursor over the entry in the etymology column will display" should be avoided; they present an accessibility barrier. How will that work for people using our mobile site, on touch-screen devices? For users of audio interfaces such as screen readers for people who are blind? For people looking at printed copies of our pages, or at our content reused on other websites? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:49, 2 September 2013 (UTC)

Should it be "Queen [[Elizabeth II]]" or "[[Elizabeth II|Queen Elizabeth II]]"?

During the past few days, I've had a rather strange disagreement with Miesianiacal over linking. I maintain that, since the word "Queen" acts as part of the person's name and is thus capitalized, it is stylistically better to link to "Queen Elizabeth II" than to "Queen Elizabeth II". This is already an accepted convention; you will have a very hard time finding a reference to "Prince Charles" or to "Queen Anne", while it is perfectly normal to link to "Prince Charles" and "Queen Anne". Featured articles such as the one about Elizabeth II link to "King George V and Queen Mary", not to "King George V and Queen Mary". It is also common and natural to link to "Sir Walter Scott" rather than to "Sir Walter Scott". I've seen FAC reviewers recommend this form of linking, as it's "easier on the eyes", but Miesianiacal vehemently opposes it for no apparent reason at all. He claims that "unnecessary pipes increase article size unnecessarily", even though the article in question has no size problems whatsoever. I tried to appease him by avoiding piping and linking to [[Queen Elizabeth II]] (a redirect) instead. He reverted that too, saying that redirects should be avoided, despite my pointing out that MoS advises against such actions.

There is also the issue of consistency within articles, and often within one single sentence. See this edit, where I attempted to change "King George VI and Queen Elizabeth" to "King George VI and Queen Elizabeth", only to be reverted with an [attempt of] explanation that such an improvement is "absolutely and entirely unnecessary".

So, what is the preferred/recommended practice regarding linking and piping? Surtsicna (talk) 18:47, 9 September 2013 (UTC)

I'm not sure the guideline has any formal objection to piped links, so long as they don't surprise the reader, which hardly applies in this case. I agree that the fully blued version looks better and that royal and other titles are usually thought of as being part of the name. It certainly seems a bit odd to unilaterally take out such pipes and then edit war over it when they're put back. N-HH talk/edits 20:59, 9 September 2013 (UTC)

(edit conflict) Please don't misrepresent me to others.

A title is exactly that: a title; it is quite apart from a person's name. Regardless, whether or not a title is a part of a person's name is really a red herring here; "Queen Elizabeth II" reads no differently to "Queen Elizabeth II". As such, the pipe needed for "[[Elizabeth II|Queen Elizabeth II]]" and the redirect caused by "[[Queen Elizabeth II]]" are entirely unnecessary; though, those for Queen Elizabeth and the Duke of York obviously are. But, whether it's "Queen Elizabeth" or "Queen Elizabeth" is irrelevant; either works just fine.

There is no guideline or even convention around this and WP:NOTBROKEN doesn't support your stance, since, as just explained, there's no actual need to use a redirect at all and avoiding one by linking directly to Elizabeth II (or George VI, or Edward VIII, etc.) does not contravene even one of the given reasons not to change redirects. Creating needless pipes does indeed needlessly increase article size. That may not matter on some small articles, but it does on those that are close to or over the recommended limit, such as Monarchy of Canada.

There's simply zero purpose in piping or redirecting to Elizabeth II, George VI, Edward VIII, George V, or Edward VII when those names are immediately preceeded by the word "king" or "queen". --Ħ MIESIANIACAL 21:02, 9 September 2013 (UTC)

How exactly have I misrepresented you? No, it does not read the same. "Queen Elizabeth II" reads much easier than "Queen Elizabeth II", and you have absolutely no argument against what is widespread and recommended practice. There is no need to use a redirect when we pipe links, but you staunchly oppose piping. Honestly, your claim that we should avoid piping because it increases article size is completely ridiculous and nonsensical. I had never seen anyone argue that, and I probably never will again. In fact, I do not believe that you yourself believe that. Three users have already disagreed with you, so it would be better if you started reasoning and finally arguing the merits or lack thereof of these changes rather than simply repeating that this is all unnecessary. It is not. Style is important; that is why we have all these pages called Manual of Style. Surtsicna (talk) 21:28, 9 September 2013 (UTC)

They read exactly the same; in both, "Elizabeth II" follows "Queen".

Can you show where your preferred method is recommended? Even if it is widespread (which hasn't been affirmed), other stuff exists.

How ridiculous you think my argument is is your opinion; it doesn't prove me wrong.

You have not given a sufficient explanation as to why creating pipes or redirects is necessary. "I like when 'king' or 'queen' are blue" doesn't quite do it. --Ħ MIESIANIACAL 21:34, 9 September 2013 (UTC)

"queen Elizabeth II" reads the same as "Queen Elizabeth II", yet one is preferred over the other. That is what "style" means. For one, "my preferred method" was recommended by N-HH above, though you chose to ignore that reply and pretend that I am the only one who disagrees with you. The fact that other stuff exists does not entitle you to revert improvements to articles, no matter how minor. The notion that it does is almost as absurd as your claim that piping should be avoided because of article size. Such weak and senseless arguments do prove that you are wrong, as you evidently have none more to offer. Surtsicna (talk) 21:55, 9 September 2013 (UTC)

No, those two do not read the same; one is grammatically incorrect and the other is not.

You see, the argument to be made here is not why I'm "entitled" to revert "improvements", but, rather, since you're the one trying to make changes to the status-quo in a number of articles, how it is that your changes are actually an improvement. Simply saying they are is not even an argument. Saying the same way has been used elsewhere doesn't prove using that way is an improvement.

The closest you have to an argument is that it reads better when both title and name are blue. But, still, that's an entirely subjective opinion. I don't see that it reads any differently at all; no differently to the blue word "Queen" coming immediately after a black word ("...he congratulated Queen Elizabeth II" versus "...he congratulated Queen Elizabeth II"). It's just a pipe or a redirect made for no reason at all. Perhaps the majority of people who weigh in here will agree with you and pipes and redirects will be the way we're all to go. And perhaps that majority won't. --Ħ MIESIANIACAL 23:09, 9 September 2013 (UTC)

No, capitalization has little to nothing to do with grammar. "queen Elizabeth", "Queen Elizabeth", "queen elizabeth", "QUEEN ELIZABETH" and "Queen elizabeth" are all equally correct when it comes to grammar. Stylistically, they are not equally appropriate. You may claim that you do not see the difference, but I do, and I am not the only one. The fact that you will only encounter references to "Prince Charles" and never to "Prince Charles" means that there are many people who do see the difference. If you really don't see it, it means that you don't think it's a step back. If you don't think it's a step back, why do you so adamantly refuse to allow those who see the difference to do what they think is an improvement? Surtsicna (talk) 23:23, 9 September 2013 (UTC)

"[C]apitalization has little to do with grammar..." Oh dear.

The "fact" of "Prince Charles" and "Prince Charles" is not a fact and the question of which to use was already addressed above via the example of "Queen Elizabeth" versus "Queen Elizabeth". The whole question is a red herring, however; the examples are not comparable: One is forced to pipe or redirect to Queen Elizabeth The Queen Mother or Charles, Prince of Wales, if one wishes to use and link the terms "Queen Elizabeth" or "Prince Charles", whereas one is not forced to pipe or redirect to Elizabeth II, George VI, Edward VIII, etc. if one wishes to use and link the terms "Queen Elizabeth II", "King George VI", "King Edward VIII", etc.

So, you're still left with "I think it looks better" as your only justification for piping and redirecting in contexts wherein it is neither practical nor necessary. Some agree with you. Didn't you notice that others don't? --Ħ MIESIANIACAL 17:32, 11 September 2013 (UTC)

"Oh dear"? I suggest reading the definition of grammar and the lead paragraph in order to grasp the difference between grammar and orthography.

Addressed how? Of course they are comparable! I am not forcing you to pipe. I was the one doing all the work, trying to bring some consistency into the article, and you reverted because my efforts were unnecessary though not detrimental. How does that make any sense? Once again, if you don't see any difference, then why do you care? Are you afraid that you yourself will be required to make the excruciating effort of piping links? This is the first time in my 5 years here that I see a user arguing against an edit by claiming not that the edit was detrimental, but that it was just "unnecessary". Surtsicna (talk) 18:29, 11 September 2013 (UTC)

Addressed in the way I said in the post above your last that it had been addressed and they are not comparable for the reasons I gave in that same post.

"Queen Elizabeth II" and "Queen Elizabeth" aren't consistent with "Prime Minister William Lyon Mackenzie King", "Governor General Vincent Massey", "President William Howard Taft", or "Archbishop Randall Davidson". Do you expect those latter to be "Prime Minister William Lyon Mackenzie King", "Governor General Vincent Massey", "President William Howard Taft", and "Archbishop Randall Davidson", and all other instances wherein a name follows an associated title to be like? I don't think its reasonable to expect to create the consistency you want.

In the course of editing, I remove needless pipes. A lot. --Ħ MIESIANIACAL 18:45, 13 September 2013 (UTC)

Doesn't it make more sense for "Queen Elizabeth II" and "Queen Elizabeth" to be consistent with each other than with prime ministers, governors, presidents and clergy? Doesn't it make more sense for them to be consistent with peers and with "Queen Victoria"? Of course it does. In the course of editing, I strive to make articles internally consistent. And I do that a lot. If that means introducing the diabolical pipes, so be it. Surtsicna (talk) 12:05, 20 September 2013 (UTC)

Reading is not how I would interpret this. Her role is queen and so it should not be piped with her name. We would not pipe [[Barack Obama|President Barack Obama]] under any circumstance, why would we do so to Elizabeth II? Walter Görlitz (talk) 21:55, 9 September 2013 (UTC)

Her role is "queen" - lower case. When referred to by title, she is called "Queen Elizabeth II". The word "queen" is capitalized because it acts as part of her name. Sir Robert Walpole is not mentioned as "Sir Robert Walpole". See also the List of Prime Ministers of the United Kingdom, which is a featured list. You'll also see Sir Henry Campbell-Bannerman and Sir Winston Churchill, rather than "Sir Henry Campbell-Bannerman" and "Sir Winston Churchill". I'm also sure you're much more likely to find a reference to "President Obama" than to "President Obama"; in fact, I believe he is normally called either Barack Obama or President Obama, though I may be wrong about that. Surtsicna (talk) 22:21, 9 September 2013 (UTC)

LOL..There goes one of them quaint editor who dreams day and night, night and day, day in day out of zee Monarchy, zee Queen and zee Commonwealth Realms. You're in for a bumpy (royal) ride. I pass.--Lubiesque (talk) 23:05, 9 September 2013 (UTC)

I support the shortest and simplest possible linking, thus exclude "Queen", "Sir", and similar titles from the linked term. -- Michael Bednarek (talk) 02:38, 10 September 2013 (UTC)

• Redirects are cheap. Use the form that suits the prose best, and there is usually a redirect. There's no need to bother with the piping. Makes it easier to read in edit mode too. :-) -- Ohc ^¡digame!_{¿que pasa?} 03:07, 10 September 2013 (UTC)

Actually, that's a point. In this case at least, it's not a piping questions at all nor is it a question of the extra words involved in piping – the option is in fact a basic one between Queen Elizabeth II and Queen Elizabeth II. As I said earlier, it hardly seems a big deal – and even less of one now – but my preference would be for the former, on simple aesthetic grounds and on the basis that "Queen", when used in this manner, does form part of a formal title and quasi-name. Btw I speak as an avowed anti-monarchist. N-HH talk/edits 09:56, 10 September 2013 (UTC)

Could it be doubted that a self-confessed anti-monarchist could offer such well reasoned advice? But (commenting tentatively as neither one nor other) it may make better sense to let links be per article title, wherever possible contextually, with unblue title as appropriate in the given context. Qexigator (talk) 11:48, 10 September 2013 (UTC)

That leads to mess such as having "Queen Victoria" and "Queen Elizabeth II" in the same paragraph, or "King George VI and Queen Elizabeth" (as in this instance). I strongly doubt that's appropriate, while "King George VI and Queen Elizabeth" looks very untidy and odd, which is why users intuitively avoid it. Surtsicna (talk) 12:06, 10 September 2013 (UTC)

All conditional upon context. Outside Wikipedia, writers, editors and sub-editors are expected to apply style rules with some discretion. Ontario has been in the news at Perth Agreement lately: perhaps a See also link there to Monarchy in Ontario (to which the above example links) would be in order? Qexigator (talk) 12:29, 10 September 2013 (UTC)

• Support including "Queen" in the wikilink It looks better that way - the other way is distracting - and "Queen" is a title, which makes it quite different from offices such as president. When linking to Lord Nelson one would not link "Nelson" but not include the "Lord" in the link. "Queen" is a title just like "Lord". Neljack (talk) 08:05, 11 September 2013 (UTC)

"Queen" and "President" are both titles. --Ħ MIESIANIACAL 15:40, 11 September 2013 (UTC)

No, "President" is an office. It is not a title in the same way that "Queen" is. Neljack (talk) 00:15, 13 September 2013 (UTC)

Given that neither is a title of Nobility nor Title of honor, that a reigning king/queen in a hereditary constitutional monarchy such as UK or Canada or any of the other Commonwealth realms is in some sense an "office-holder", that a past POTUS is by courteousy and custom of the people of the USA allowed the honorific style or form of address "President", that in practice an incumbent POTUS is commonly and customarily addressed as President of the USA as if it were a title, and so also are presidents of other sovereign states, unlike the president of, for example, the UN General Assembly or Security Council - by what rule is it deemed that President is not a title in the present context? Qexigator (talk) 08:46, 13 September 2013 (UTC)

"Queen" (or "king") is an office; the Canadian Constitution Act 1982: "An amendment to the Constitution of Canada in relation to... (a) the office of the Queen". But, that's irrelevant; the subject is titles. If "President" is not a title, what is the word "President" when it comes before the name of an individual who holds a presidential office? --Ħ MIESIANIACAL 17:23, 13 September 2013 (UTC)

Why "neither right nor functional"?[6] --Qexigator (talk) 17:02, 13 September 2013 (UTC)

I don't see how "King George VI and Queen Elizabeth" can be right and functional. First we have "King George VI", with title excluded from the link; one word later, we have "Queen Elizabeth", with title included in the link. Why is such mess desirable? Miesianiacal has not even attempted to explain why "King George VI and Queen Elizabeth" is better than "King George VI and Queen Elizabeth". I understand that he may not be willing to trouble himself with piping but what kind of a problem does he have with other people troubling themselves with such matters? When Glenmeister introduced pipes with the explanation that doing so "creates more consistent and more aesthetically pleasing, better looking links", Miesianiacal reverted with simple statement as simple as "pipe not needed" and "de-piping". Here he acknowledged that he sees no difference between "King George VI and Queen Elizabeth" and "King George VI and Queen Elizabeth", so I wonder why he dislikes when people who do see the difference act upon it. Surtsicna (talk) 17:47, 13 September 2013 (UTC)

• With the above example, I agree it's silly and ugly. We should ensure that adjacent wikilinks are visually consistent when their respective subjects are closely related. As I mentioned in an earlier post, I see no need to pipe [[George VI|King George VI]] when the redirect [[King George VI]] can be used, and it makes sense for the link to include the title – omission looks weird to these eyes. But I would see no alternative for "Queen Elizabeth", which has to be piped as the term can never be unambiguous, and using "Queen Elizabeth The Queen Mother" isn't right in that context. -- Ohc ^¡digame!_{¿que pasa?} 18:18, 13 September 2013 (UTC)

Well, Surt., your last reads like a preamble, but is there no reasoned answer to explain why "neither right nor functional"? Or is "creates more consistent and more aesthetically pleasing, better looking links" all there is? Let me add here, that, while it may seem to look better to blue up, it is better only to blue the article name. New readers will soon enough catch on that normally the blue bit is the name of the article. But where that doesn't work in a given context, a pipe may be unavoidable. Thus: where context allows, only linked article name in blue, leaving title or whatever in black. Only if that is impracticable, use pipe, but always try to write so as to avoid letting a pipe hide the name of the linked article, if that name has not already appeared in the text. Blueing up is for links, not for decorating the text, just as other devices, such as listing bullets, conventionally have a specific function, which is not decoration. Unless there is a rule against it, using, or even creating, a redirect would be a preferable device. Qexigator (talk) 18:43, 13 September 2013 (UTC)

We could discuss whether it's appropriate to force readers to catch on why we have "King George VI and Queen Elizabeth" rather than "King George VI and Queen Elizabeth" or even "King George VI and Queen Elizabeth" (and then to catch on why the articles are titled George VI and Queen Elizabeth The Queen Mother, and so on). However, for now, I would only like to point out that Miesianiacal vehemently opposes even using redirects such as King George VI. Such a redirect would enable us to have the much more sensible "King George VI and Queen Elizabeth" without piping. Surtsicna (talk) 19:11, 13 September 2013 (UTC)

Describing what you think is a mess doesn't really explain why the same is "definitely neither right nor perfectly functional". --Ħ MIESIANIACAL 18:49, 13 September 2013 (UTC)

Stating that something is unnecessary does not render that wrong, unacceptable or detrimental. Stating that you do not see any difference means that you do not consider it better or worse.

If someone else considers it better and you don't consider it worse, why oppose it? Surtsicna (talk) 19:11, 13 September 2013 (UTC)

Surt.: When a blue appears among the black, readers very quickly find out it links to another article, usually and normally of the same name. That is as simple as seeing how the contents box links to the parts of the article. What is less easy for readers to understand is the use of pipes, especially those which depart from the reader's expectation of the norm and hide the name of the linked article. Some readers (like me) may find that a nuisance, and usually regard it as an editing affectation, spoiling the enjoyment of consulting Wikipedia with its peculiar usefulness in providing cross-links which can be read and followed with ease. Qexigator (talk) 22:14, 13 September 2013 (UTC)

Okay, how about redirects instead of pipes? As I said, Miesianiacal is just as much opposed to redirects, even when they correct mess such as "King George VI and Queen Elizabeth" by replacing it with "King George VI and Queen Elizabeth". Doesn't "King George VI and Queen Elizabeth" make more sense to you than "King George VI and Queen Elizabeth"? Surtsicna (talk) 22:27, 13 September 2013 (UTC)

In my view, using redirects is a better device for bluing up, but there may be some WP discouraging creating redirects other than for the purpose of re-naming an existing article, or to allow for readers who are likely to look for a topic under one name when Wikipedia has chosen to put it under another. And I still see it as preferable as a general rule not to blue up artificially using that device or any other. Qexigator (talk) 22:42, 13 September 2013 (UTC)

There is no such policy or guideline. In fact, there is a guideline that advises users not to avoid redirects where they can he helpful. That guideline certainly applies to the case where "it's silly and ugly" (as Ohconfucius put it) to neither pipe nor redirect. If someone else considers an edit to be an improvement and you don't think the edit is detrimental, why oppose it? I am having a truly hard time trying to understand the logic there. Surtsicna (talk) 22:50, 13 September 2013 (UTC)

If there is no WP impediment, using redirects is the better option for blueing up. But it may be that here we have one of those instances, in and out of Wikipedia (one could say, in real life) when what at first seems "silly and ugly" can later be better understood as helpfully plain and simple. Qexigator (talk) 23:21, 13 September 2013 (UTC)

I dislike a "blueing up" as much if not more than the next man, but I'm also in favour of sensible and intuitive linking. If the writer wants to write simply "[[George VI]]", then there's no issue, but if the editor wants to use the terms ""King [[George VI]] and [[Queen Elizabeth The Queen Mother|Queen Elizabeth]]"", the coherence of the word King in relation to the terms is gone. Of course, that would also open up the possibility that some innocent and inexperienced editor might add/chain the link to [[King]] or [[Monarchy of the commonwealth country|King]], like they often do with [[President of the United States|President]] [[Barack Obama|Obama]], but that's a side issue.

In many cases, such as the above, the redirect is unambiguous if correctly chosen. OTOH, I've seen editors pipe links in such creative fashion that the result is a mysterious Easter egg. -- Ohc ^¡digame!_{¿que pasa?} 02:00, 14 September 2013 (UTC)

Tapping on that "Easter egg" resulted in a pleasant surprise. It goes some way to settling the question, allowing a realistic margin of discretion (and misapplication) to intuitive decisions about deviating from the norm. But where two or more devices are available, one editor may allow less weight than another to a question such as "coherence of the word King in relation to the terms" in the above comment. What prevails to change status quo? (My strength of feeling is not great either way). Qexigator (talk) 08:54, 14 September 2013 (UTC)

• Comment It seems to me we should treat "Queen" (or "King") in the same way we would treat "Prince" or "Princess". Suppose we were referring to the Queen before she succeeded to the throne. Would we wikilink to her as Princess Elizabeth or Princess Elizabeth? Surely the former. It should be the same with "Queen". Neljack (talk) 00:12, 16 September 2013 (UTC)

• The latter is what I would classify as an Easter egg that ought to be avoided, IMHO. Elizabeth is a firstname that is imprecise about a certain individual even though the context may not be, while 'Princess Elizabeth' is unambiguous. -- Ohc ^¡digame!_{¿que pasa?} 01:22, 16 September 2013 (UTC)

• Your point would also apply to monarchs without numerals e.g. Queen Victoria or King John. Neljack (talk) 01:42, 16 September 2013 (UTC)

• I'm in favour of bluing-up "the works", not just the name portion whether suffixed with Roman numerals or not. -- Ohc ^¡digame!_{¿que pasa?} 04:04, 16 September 2013 (UTC)

Linking of small countries

There has been a discussion at Wikipedia talk:WikiProject Wales#"Overlinking", prompted by User:John's systematic removal of links to the article on Wales from a number of Wales-related articles. For info - Wales has about 5% of the UK's population, 0.05% of the world's population. Most editors in or involved with Wales seem to want the links to remain; John is among those who say that they should be removed per MOS as Wales is a ""major geographical location". Views? Ghmyrtle (talk) 18:41, 25 September 2013 (UTC)

I'll briefly repeat my standard line, for what it's worth: in the guidelines as actually written, there is no absolute bar on linking "major geographical locations" – the guideline simply says they are "not usually linked" when they are not "relevant to the topic of the article". Hence, if relevant, a link is fine; even if not directly relevant, there's still no outright prohibition. Beyond that, there is also no clear widespread consensus as to what would constitute "major" anyway or to mandate the mass removal of such links.

As to the specific question here, and regardless of the obvious difficulties in defining "major", I'd agree that Wales, with due apologies, is probably not in global terms a major country or area. I can't see what potential benefit there is to widespread removal of such links, let alone when weighed against the potential loss of utility to anyone who actually might want to navigate quickly and easily to the Wales page from a related page to learn more about the place. Why make that harder when there's no need to? N-HH talk/edits 19:02, 25 September 2013 (UTC)

The argument for retaining the links was that unless we link Monmouthshire, Wales rather than Monmouthshire, Wales, the hypothetical reader will be unable to find the Wales article even though it is clearly linked at the Monmouthshire article. This is what we call "chain-linking" and is an example of what this guideline exists to try to avoid; a dilution of useful links with useless ones. The argument was also raised that as Wales receives x number of hits per day, this proves that readers do not know about it. As User:Drmies pointed out, this logic would mandate overlinking, especially to articles like United States which are very highly-read. This is clearly nonsensical. Oh, and Ghmyrtle, zero is a number too. --John (talk) 19:14, 25 September 2013 (UTC)

See wikt:a number of. Ghmyrtle (talk) 20:18, 25 September 2013 (UTC)

Ah, Wiktionary. Now there's a quality resource...--John (talk) 20:46, 25 September 2013 (UTC)

a number of things: several of a particular type of thing. Ghmyrtle (talk) 22:50, 25 September 2013 (UTC)

But, as I asked, why make it harder? And who's defining "useless" or "useful" here? These are subjective judgments that cannot be applied universally, as if they apply to all readers equally. I would agree that the number of hits on a page is an indication of how interested people are in that topic – whatever their degree of knowledge about Wales at the outset – and hence is the closest we have here to any objective evidence as to what is "useful". And as for "dilution", if we really credit readers with any intelligence, we'll accept that they can work out which links they might or might not want to use on any one page. It's rather patronising, surely, to suggest that a small group of editors should be making that decision for them in advance. N-HH talk/edits 19:31, 25 September 2013 (UTC)

"We're not", "the Wikipedia community", and "No, it's considered more patronising to chain-link", in that order. --John (talk) 19:33, 25 September 2013 (UTC)

In that order too:

• Yes you are – your suggestion is to add an extra stage, so that people who might want to open the Wales page have to click the Monmouthshire link first and then access "Wales" from that page. Under what possible definition is that not making the process more difficult?
• Where has the "Wikipedia community" – rather than one or two editors who constantly, regardless of page context, run these scripts and who regularly run into objections, as they have here – decided what kind of thing is more or less useful in the wider sense when it comes to linking, let alone in this specific context of links to Wales?
• Why is it "considered more patronising to chain-link"? By whom? In what way exactly? N-HH talk/edits 11:54, 26 September 2013 (UTC)

N-HH talk/edits 11:54, 26 September 2013 (UTC)

• I'll repeat my standard line: I didn't look at John's edits, I just responded to the main point in that thread and argued against chain-linking (I didn't know that term at the time). As the case is presented here, I agree with John. Where this "small group of editors" concept comes from I don't know (I assume that the overlinking guidelines were not concocted in anyone's basement), but if this talk page reaches a bigger group of editors, that's fine with me. Drmies (talk) 20:16, 25 September 2013 (UTC)

The discussion Wikipedia talk:WikiProject Wales#"Overlinking" isn't limited to geographical chains such as "Monmouthshire, Wales" (which may be extended in either direction; whose two linked terms are actually separated by plain "south-east" in the original illustration).

That original illustration[7] includes script-assisted replacement of "Belgian slate" (with a piped link to Belgium) by "Belgian slate" (no link). Beside the small country argument, context ("Welsh slate" is linked in the same sentence) suggests that examination and improvement of the Belgium link is appropriate, rather than its elimination.

Script-assisted unlinking of Wales or Belgium when piped to "Welsh" or "Belgian", or "Welsh X" or "Belgian X", might help more than it hurts --by prompting someone who watches to link a superior target in its place-- or might hurt more than it helps.

The latter dilemma, if real, may pertain to the name of a big country whose adjective is entirely regular, eg Russia piped to "Russian" or "Russian X". --P64 (talk) 01:13, 26 September 2013 (UTC)

Chain-linking and bunch-linking degrades the linking system. Neither should be used unless there is a compelling reason to do so. I don't find a compelling reason to link "Wales" when it's right next to Monmouth, in which a link to "Wales" occurs prominently in context. We encourage readers to go first to the more specific link. Tony (talk) 06:03, 26 September 2013 (UTC)

So we acknowledge that people might wish to open a link to "Wales" from the first page, but we're saying that we are going to make them take the long way round and open an intermediate page first to get there? And who's to say that the script won't next alight on the Monmouthshire page and remove the link to Wales from there as well anyway? I appreciate the aversion to the bunching that comes with chain-linking, but that's a slightly different issue from the "major geographic features and locations" point at WP:OVERLINK. Either way, it seems that we are focusing on the aesthetic preferences of a few individual editors rather than the broader interests of what is, after all, a pretty large and diverse readership. N-HH talk/edits 12:48, 26 September 2013 (UTC)

Linking same information multiple times in infobox

The issue of linking the same information (e.g., year, team), multiple times within an infobox has arisen here.--Epeefleche (talk) 01:34, 5 October 2013 (UTC)