User talk:Jimbo Wales

(Manual archive list)

How much should we trust Wikipedia? (again)

This is in the news today, and is another story about the old chestnut of WP:NOHOAXES. The information was added here by a Cambridge University IP address in November 2007, and removed here in October 2010. I suppose we could all beat our breasts at our failure to find this sooner, but the blame rests with the people who do this sort of thing.--♦IanMacM♦ ^{(talk to me)} 17:26, 16 April 2015 (UTC)

• And HERE is another similar-but-different bit from the Washington Post about the Jar'Edo Wens hoax and a recent breaching experiment by Greg Kohs testing how thoroughly hoaxy hoaxes are caught. Kohs's WP:BEANS revelation is that edits with footnotes have a tendency to stick, even if the footnotes don't support anything. I will also note that each and every one of Kohs's hoaxes were made possible by the fact that IP editing is greenlighted here instead of some sort of real name registration and sign-in-to-edit policy, like the big kid websites all use. Carrite (talk) 20:35, 16 April 2015 (UTC)
• Well if it meant in an emergency looking up the problem on Wikipedia or calling 911 id do the latter. - Knowledgekid87 (talk) 20:41, 16 April 2015 (UTC)
• I'm surprised there's no anti-hoax Wikiproject or task force on Wikipedia. It seems like something that really needs a lot more scrutiny. Bosstopher (talk) 21:25, 16 April 2015 (UTC)

How would an anti-hoax Wikiproject know which articles needed 'scrutiny'? AndyTheGrump (talk) 22:18, 16 April 2015 (UTC)

By planting hoaxes themselves and monitoring how fast they get removed. Count Iblis (talk) 22:48, 16 April 2015 (UTC)

I think an easy way to start would be through adding ISBN codes to every single book cited, to ensure that the books in question actually exist. A bot could be created to scan cite templates that contain an ISBN code, see if the code matches the book cited. Another possibility (tbh I'm not sure if this is possible, I can't code for shit), would be a bot that looks up books/journal articles cited on google books/jstor and searches to see if the title of the wikipedia article appears in the source (if a bot like this existed it would have picked up the Colin Hilton hoax). These two bots could flag articles for examination by the wikiproject. Another task could be collaborating to review articles by known/suspected hoaxers. User:KuchenZimjah, who's had a lot of his created articles deleted as hoaxes, also seems to have created a few legitimate articles. But as there was no concerted collaborative effort to discern which of his articles are hoaxes, there's a chance some of his hoaxing is still on the encyclopedia. It would be useful to have a hub where things like this could be evaluated. Bosstopher (talk) 22:51, 16 April 2015 (UTC)

An ISBN check was how the Bichoilm conflict hoax was caught. One-off pranks like Jare'Edo Wens are embarrassing but not as big of a deal as some make them out to be. Non-obvious fakes like Bichoilm are more damaging and need these kinds of verification efforts, especially at GA and FA. Gamaliel (talk) 23:13, 16 April 2015 (UTC)

You are aware, of course, that ISBN numbers are nothing more than UPC numbers for books, cash register and accounting selling numbers for publishers and retailers? And that paperback and hardcover editions of the same book have different numbers? And that every publisher's edition of the same title has its own number? And that millions of books from the past and present do not have, and never will have, ISBN numbers? This doesn't get into the unreadable, ugly mess that reading lists become with ISBNs. Or the fact that they are essentially useless in searching for books online and in libraries, since author + title + subject are the parameters generally used. ISBNs are useless... Carrite (talk) 17:31, 17 April 2015 (UTC)

I admittedly know very little about bot programming and ISBN codes. My suggestion of giving every citation an ISBN was a bit of a stupid one. But I have noticed when working on articles using the ISBN code to autocomplete the reference template always produces something vaguely resembling the title of the book in question. If the editing UI's citation tool already allows for book titles to be produced, I dont see why the string produced cant be compared to the book title listed on wikipedia to see if they roughly match. As already mentioned hoaxes like the Bicholim conflict have already been detected this way. Bosstopher (talk) 23:49, 17 April 2015 (UTC)

See Wikipedia:WikiProject Fact and Reference Check.—Wavelength (talk) 00:36, 17 April 2015 (UTC)

• I'm surprised no Wikipedians have jumped in yet to belittle or disparage the qualifications or opinions of the Telegraph or Washington Post journalists who wrote those stories. I remember one time when Jimbo said of a columnist who criticized Wikipedia that he was "not a real journalist." Is the tide turning on WP about being more accepting of outside criticism? Cla68 (talk) 23:04, 16 April 2015 (UTC)

• While embarrassing, I don't see the hoaxes as being the bigger problem. What is more concerning is the number of times I check the source for a surprising, especially damning, or especially praise-ful statement in an article, and find that the source says no such thing. I've spent the last few days edit warring with a newly established SPA account that keeps adding statements to the Atorvastatin article that the drug causes Alzheimer's disease. What the cited sources actually say is that in a very small portion of users, it causes mild cognitive problems that stop within days of discontinuing the drug. Formerly 98 ^{talk|contribs|COI statement} 23:52, 16 April 2015 (UTC)

FYI, this search [1] suggests there's a bit of "noise" amongst the health blogs about Lipitor & Alzheimer's. The editor in question may be a caregiver or loved-one of someone with Alzheimer's who is trying to add WP:TRUTH to the article. Good luck. JoeSperrazza (talk) 00:01, 17 April 2015 (UTC)

• There's a a lot of good ideas in this thread that deserve followup. Wikipedians could benefit from having in-house tools to verify, fact check, and improve the quality of our content. It would be great if the WMF could focus on developing these important tools. Viriditas (talk) 01:24, 17 April 2015 (UTC)

A journalist from a reliable source quoted the claim about the invention of the butterfly stroke here. Journalists have a love-hate relationship with Wikipedia, criticising it one minute while using it as their crib sheet the next. The problem is that a well-meaning but uncited statement could be a deliberate hoax. Since Wikipedians give their time free of charge, they cannot spend 24/7 checking often obscure articles for factual errors, deliberately introduced or otherwise.--♦IanMacM♦ ^{(talk to me)} 06:18, 17 April 2015 (UTC)

Which is exactly why we would benefit from tools that would help us complete the task. I've never understand this strange argument that because we aren't getting paid we shouldn't be concerned with quality. Anyone using tools to patrol for recent changes, to navigate categories, and to review new articles, would be in a position to use these tools. Yes, we're all volunteers, which in fact has the opposite implication you intended. Since Wikipedians do this for the love of it, we are in a better position to spend our time, leisurely if need be since there's no deadline, looking for errors. Viriditas (talk) 12:16, 17 April 2015 (UTC)

Indeed. There is a one-month backlog in patrolling new pages. We have tools that aggregate these pages for review, but the reviewing itself is mostly manual. Major development efforts (VisualEditor) focus on making it easier for new editors, on the premise that we have a shortage of new editors creating content, because in theory they have trouble doing that. Reality is that Category:Pending AfC submissions has a backlog, so the shortage is in patrol volunteers. Tools with Watson-like intelligence would be nice: "Article A says 'fact x', but article B says 'fact y', which contradicts 'fact x'. Which is correct?" Wbm1058 (talk) 14:36, 17 April 2015 (UTC)

An interesting tool would be to empower readers who don't want to go through the effort of editing an article, still have a chance to provide quick feedback in a graphical way. For example, have a one-click "highlighter" tool, which the reader can then highlight a word, a sentence, or a paragraph, and then the interface would prompt them to either select "green" for "good", or "red" for "bad". None of this highlighting would appear on the live page, but a tab at the top (similar to "edit" and "new section") might say "heat map". Then bona fide editors could have a look at the heat map to see if there are any reader-induced concerns about certain spots within the article. P.S. I'm a banned editor, so if someone feels the need to remove this comment to "punish" me, then restore it under their own account, my feelings won't be hurt. - 2001:558:1400:10:CC0F:8EAD:7C36:144A (talk) 14:45, 17 April 2015 (UTC)

WMF was playing with "article feedback" buttons (a disguised mechanism to get people to make their first edit, I think, but that's neither here nor there...) What about the idea of a simple "Content Flag" button, in which a button is clicked, opening a screen in which a simple alert message can be typed — "There is vandalism in this article in the third section" or whatever. People who are intimidated about changing content might be willing to report bad content. More eyes watching means fewer hoaxes. These alert messages could be run on a single page queue, like the late unlamented Feedback responses were. Carrite (talk) 17:22, 17 April 2015

When I review an article, I almost always review the talk page. Just the other day, I removed an image from an infobox because of comments on the talk page. For me, at least, the question is how do we highlight and prioritize important comments about article content from the corresponding talk page? Your idea about a pushing this stuff to a queue is great. I'm also thinking that allowing editors to categorize talk page comments based on article elements imported from the current version (image, infobox, lead, etc.) could push these concerns to more individualized queues. For example, if there was a problem with the rendering of the table, we have editors who specialize in creating and maintaining tables. If there's an image licensing problem, it could go to the appropriate board. Of course, we already have this framework in place with the appropriate categories, but we can't deliver the information and present it a way that will allow more editors to participate in problem solving. All of this sounds like a job for subscription-based notifications, and I've been saying this before echo was ever designed. If you look closely at user contributions, you'll find that most people find a niche and specialize in it. What we want to do is deliver the problematic content to our content "experts" and let them fix the problems. That's always been the greatest disconnect here. Connecting editors with other editors is one thing, but connecting them with content they can help fix is another. Traditionally, admins might go on such hunting expeditions, but except for copyvio, it rarely has anything to do with content. I think the way forward is to simply take the existing content maintenance categories and feed them into the notifications. At that point, you can appropriately gamify the user experience and encourage people to fix accuracy problems for fun. Viriditas (talk) 21:15, 17 April 2015 (UTC)

• Don't be silly. What is far more important is developing tools for IPs to edit via mobile, and via the visual piece of shit editor, so that even more people can vandalise the encyclopedia. You didn't expect the WMF to do anything functionally useful, did you? Let's face it, that would be a first. Black Kite (talk) 21:55, 17 April 2015 (UTC)

Haven't done any controlled studies, but anecdotally it seems like a disproportionate percentage of my recent reversions are mobile IP edits, and it seems to be getting worse. Wbm1058 (talk) 02:19, 18 April 2015 (UTC)

I've actually noticed a number of mobile vandal edits that were reverted on a lot of pages.--Mark Miller (talk) 04:44, 19 April 2015 (UTC)

Putting accuracy in perspective

Given the large number of non-sourced articles written from editors' "personal knowledge", it wouldn't surprise me to find more inaccuracies introduced by good-faith edits than by intentional vandalism. Ultimately whether the inaccuracy is good- or bad-faith isn't important. Correcting it is what's important. A chapter titled "in screen we trust" in ISBN 978-0385539005 says that nearly 25 percent of consumer credit reports contain errors, data brokers such as Acxiom admit that 30 percent of the data they maintain may be inaccurate, tens of millions of electronic medical records contain incorrect information about patients, and British police officials admitted that more than twenty thousand people had been wrongly branded as criminals due to Criminal Records Bureau errors. Wbm1058 (talk) 15:11, 17 April 2015 (UTC)

[[Citation needed]] Liz ^{Read! Talk!} 22:43, 17 April 2015 (UTC)

Cool. So Wikipedia's unreliability isn't a problem. Relax everyone. No need for change. The model is just fine. Carry on as you were. --Anthonyhcole (talk · contribs · email) 23:18, 17 April 2015 (UTC)

Just to be clear, if I see it on Facebook I can add it to Wikipedia right?Camelbinky (talk) 00:22, 18 April 2015 (UTC)

I'm puzzled by these responses. I did provide a citation. Facebook is not a reliable source. Wikipedia's unreliability is a problem. And other datasets you would expect to be highly reliable aren't either, so Wikipedia is not an outlier in that regard. Especially since we use some of those data sources as "reliable sources". Society as a whole perhaps doesn't value accuracy as much as it should, and indeed Wikipedia editors should strive for a higher level of accuracy. On the other hand, it's not possible to perfectly know everything (like next week's weather or the exact date someone was born), and in those cases we settle for the best approximation. Wbm1058 (talk) 01:44, 18 April 2015 (UTC)

This myth, that Wikipedia is reasonably reliable, is a myth, and will remain a myth until Jimmy's foundation actually commits some funds to actually determining the actual (not mythological) reliability of the faux knowledge Wikipedia pushes. That they haven't done so yet strikes me as deliberate ignorance. Like they're scared of what they'll find. Jimmy and his foundation - at least until recently - have only cared about the myth and keeping it alive.

I say "until recently", because the recently released State of the Wikimedia Foundation 2015 Call to Action [2] lists a set of objectives. One of the items under the heading "Focus on knowledge & community" is "Improve our measures of community health and content quality, and fund effective community and content initiatives."

Five years ago the Wikimedia Movement Strategic Plan [3] resolved to measure and measurably improve the quality of our offering, and no resources were allocated and it did not happen. Hopefully, this time round, something will actually happen. --Anthonyhcole (talk · contribs · email) 10:29, 18 April 2015 (UTC)

They love to spend donation money on engineers and funding regional user groups of dubious efficacy. They don't like funding statisticians and social scientists to actually take a look at who and what they've got in terms of a volunteer base and actual content and how to help make it better. Carrite (talk) 01:37, 19 April 2015 (UTC)

They should spend more money on more good engineers - and they seem to be doing just that. But it's not an either/or thing. As for regional user groups - some of the best initiatives have been germinated and nurtured in WMUK and WMDE. It's good to have well-resourced, decentralised idea-hot houses, and the chapter model serves that function up to a point. There may be a better model but I haven't seen it yet. I hope they don't flush away the chapters before developing a proven better model. --Anthonyhcole (talk · contribs · email) 02:25, 19 April 2015 (UTC)

Putting coverage and perspective in perspective

To me, a far worse problem, than reliability of occasional incorrect or falsified data, is the lack of overall coverage about various topics. Of course it is important to nitpick the specific details when checking the contents of a page, but in general, the "elephant in the room" for many articles is the lack of overview summary about each topic, and the related lack of key details about major facets of a topic. Many pages read like a "shaggy dog story" with random, disorganized facts plus questionable quotes about a topic. Because Wikipedia is an encyclopedia, then each page should provide encyclopedic coverage about a topic, such as addressing the "Six Ws" (who, what, when, where, why, and how) which could better explain the topic across a spectrum of facets. For example, when entertainer Michael Jackson died in June 2009, the biopage did not discuss Jackson as a film actor (The Wiz or acting in music videos), and in fact some people removed the term "actor" from Michael Jackson's page in trying to re-force a more narrow view. On balance, if a page focuses on the wider, overall aspects of a topic, then the specific details become far less important in the mix of numerous facts about the topic. Also, in general, avoid including quotes within a page because that complicates the text for exact-wording, copyright restrictions, or elevates problems of wp:POV-pushing the remarks (or rants) of individual people above the overall text about a topic. To avoid false details, avoid too much emphasis on details except when specifically sourced to explain crucial facts about the topic. When wp:NPP patrolling for new edits, then consider removing any direct quotes which might add little overall perspective but could complicate the fact-checking and NPOV-balance of the text. -Wikid77 (talk) 19:21, 18 April 2015 (UTC)

Science and technology quiz

Editors can test their knowledge of science and technology at http://www.pewresearch.org/quiz/science-knowledge/.
—Wavelength (talk) 23:18, 17 April 2015 (UTC)

That was fun. Perfect score.Camelbinky (talk) 00:21, 18 April 2015 (UTC)

Yup. Not hard. AndyTheGrump (talk) 00:34, 18 April 2015 (UTC)

Same. Sad part is, at least at the time I took the quiz, that put us in a measly 7% of respondents that got every answer correct. Given how easy that was I find that rather depressing. Capeo (talk) 01:27, 18 April 2015 (UTC)

"You scored better than 93% of the public and the same as 7%." I even correctly guessed which was the hardest question. So Sue G. is right: Wikipedians are very smart. Wbm1058 (talk) 01:30, 18 April 2015 (UTC)

The last couple of questions were tough. I clicked the answer saying I was male, but they never told me whether that was correct or not. Short Brigade Harvester Boris (talk) 02:03, 18 April 2015 (UTC)

Yeah, I got the first 13 but never got confirmation that my answers about my age or education level were right. The Blade of the Northern Lights (話して下さい) 04:47, 18 April 2015 (UTC)

We could use something like this as a WP:COMPETENCE test; add some logic puzzles, a reading comprehension test and a few grammar questions to cover most of the bases. It's time we killed off the "anyone can edit" pie-in-the-sky idealistic nonsense. Roger (Dodger67) (talk) 06:33, 18 April 2015 (UTC)

A logic puzzle such as this one, perhaps? BethNaught (talk) 06:52, 18 April 2015 (UTC)

While I did "brag" that I got a perfect score, we as Wikipedians should remember- not everyone's job requires even that quiz's basic science and technology knowledge. If you replace light bulbs at a nuclear plant and make $15 to $18 an hour, you're probably just as happy not knowing covalent bonds or the three colors of quarks, as an adjuct professor who probably makes the same as that when you count up all the hours truly "worked" whether in a classroom, office hours, or other required time at home. I never required any the knowledge of any thing on that test for any of my businesses, nor for my college degrees. Knowledge for knowledge's sake. Yes, everyone probably should read at least some thing like Discover or Scientific American to at least get the broad view of science, an appreciation of space travel, and a general view of the state of our climate and what is/isn't being done. Oh, and that evolution is real, now that's a question to use on a real litmus test of "qualified editor" for Wikipedia. An 90 IQ editor who can research and add correct citations can, and probably often does, do a better job than a 145 IQ editor who "knows what's right" and adds and edit wars because (s)he knows "better" than others, including apparently the published material. We should see if we can get enough editors to submit to an IQ test to get some data on this and compare edits to IQ. I have a feeling some "average", and statistically below average by a standard deviation or even two, would show that they are actually slightly better or even in editing. And if the average Wikipedian's IQ is statistically skewed to be higher by a standard deviation from the average (which I highly, HIGHLY doubt) we can adjust our data accordingly and I think my premise would stand.Camelbinky (talk) 12:59, 18 April 2015 (UTC)

Camelbinky, your mention of evolution leads me to refer you to this set of questions.

• The Origin of Life—Five Questions Worth Asking

—Wavelength (talk) 19:43, 19 April 2015 (UTC)

Your comment is laughable and I sincerely hope you don't edit on any science based articles and specifically regarding evolution. Such fringe ideas MUST per our policy be only used to describe FRINGE ideas and not presented as scientific fact, you're very lucky this is a talk page, otherwise your use of posting that horrible "source" would have been deleted from an article. Keep your religious beliefs to yourself please, this isn't a church. Churches have walls for a reason.Camelbinky (talk) 21:15, 20 April 2015 (UTC)

If you have found any factual or logical error(s) in that publication, then please kindly identify it (or them).

—Wavelength (talk) 04:03, 22 April 2015 (UTC)

BethNaught, that puzzle ("Cheryl's Birthday") reminds me of the "Impossible Puzzle".

—Wavelength (talk) 04:29, 19 April 2015 (UTC) and 13:41, 19 April 2015 (UTC)

I scored "We're very sorry, but the page could not be loaded properly. This should be fixed very soon, and we apologize for any inconvenience." They clearly need some clever science people to make things work properly :-) Squinge (talk) 08:20, 18 April 2015 (UTC)

Working now! 13/13 I wish quiz nights in the pub were this easy. I was wondering how many of those questions I'd have got right when I was a student. ϢereSpielChequers 21:35, 18 April 2015 (UTC)

Talk:John Prescott is indef semi-protected

Hi, did you see the note that I left at Talk:John Prescott#'box of kink' vandalism? --Redrose64 (talk) 12:45, 18 April 2015 (UTC)

Please just unprotect. Jimbo no longer performs admin actions. 156.61.250.250 (talk) 13:45, 19 April 2015 (UTC)

@156.61.250.250: You were saying? --Redrose64 (talk) 23:05, 20 April 2015 (UTC)

Thanks, Jimbo. 156.61.250.250 (talk) 18:19, 21 April 2015 (UTC)

Controlling developing world development..

A few months back I commented somewhere about what wikipedia should be doing for the future on here and was responded to by Erik Moller and a few others who liked my ideas. One of the biggest problems we're facing in the next five years I think is as many parts of the developing world come online we're going to increasingly see poor quality ip editing from around the world. India (with the exception of the elite fluent editors on here) and Pakistan as you might know have had some of the poorest quality articles on here for sometime, given the higher traffic and that most of the articles are off the English speaking world radar, particularly rural localities. This has now spread to Bangladesh. Look at Rangpur City and Government Laboratory High School, Comilla for instance. I once cleaned up some towns in an Indian state and they've since largely gone back to their original mess, with lists of "famous" locals, POV, capital letters etc. It's going to get out of control for most developing nations as they come on line and we get more traffic in areas off the radar. One of my suggestions was the idea of creating wiki ambassadors to facilitate development between wikipedias, but also the idea of setting up training schemes within such countries themselves and encourage local governments and schools to get involved with the project and try to maintain some level of quality. For countries like Pakistan and Bangladesh we need wiki ambassadors and editors consistently working to monitor progress and maintain quality and to nurture people into becoming good editors with a better command of English and how to reference and write articles. I think we're going to need to organize cleanup drives for such countries as traffic increases. It's something I'd like the foundation to seriously consider. We're going to see a rapidly increasing surge of traffic from all over, Latin America, Africa and Asia from mobiles and this is really going to be a major problem in maintaining quality in articles in developing world nations. If we don't start planning something now most of the articles in the developing world on here are going to degrade in quality and it'll get out of control.♦ Dr. Blofeld 14:17, 19 April 2015 (UTC)

By avoiding exposure to common errors, we who have English as a native language can more easily avoid committing them, and thus we can contribute to setting a good example for foreigners. Avoiding such exposure completely is practically impossible, because common errors are sometimes made in mass media productions and in scholarly publications.

• Common Errors in English Usage and Grammar Check

—Wavelength (talk) 19:37, 19 April 2015 (UTC) and 22:09, 19 April 2015 (UTC)

There's a massive difference between the odd typo or common factual/grammar error and a typical article on a town in Pakistan for instance with extremely poor prose, excessive unsourced POV, and non notable lists of schools in CAPLOCKS etc!♦ Dr. Blofeld 22:18, 19 April 2015 (UTC)

Please illustrate the massively bigger problem by providing permanent links.

—Wavelength (talk) 23:52, 19 April 2015 (UTC) and 23:57, 19 April 2015 (UTC)

It doesn't take fifteen people funded by a massive research grant to illustrate Dr. B's point... Just watch THE QUEUE for a couple days. India is working the patrollers really hard, shall we say. Carrite (talk) 04:35, 20 April 2015 (UTC)

This gem for instance, unsourced yet a bare url link to Facebook...♦ Dr. Blofeld 11:15, 20 April 2015 (UTC)

India does deserve to be watched, but I've been saying for years- watch China. Not only is it state-sponsored but you've got that fresh off the lot superpower smell still in the air and it has whipped up a lot of pent up "fix what the Westerners got wrong" mentality. You've got excessive jingoism and chauvinism (real meaning, not the type of males bashing women), for instance where China has forever and always rightfully controlled Tibet is often pushed based on dubious claims and Chinese sources as "being more accurate than pro-Western biased sources".Camelbinky (talk) 08:04, 20 April 2015 (UTC)

Using User:AlexNewArtBot, I've been looking through many of the recent articles on India and Pakistan. The main problem seems to be brevity with lots of one-liner stubs. Maybe an easy way to intensify control would be to try to find three or four volunteers who would be prepared to look through the new additions and perhaps offer advice to the most frequent contributors.--Ipigott (talk) 08:56, 20 April 2015 (UTC)

It's also a major problem with older articles, with frequent BLP violations (adding names of living people), using the article for promotion of a school, businesss, etc, and copyright violation. I've had to give up those articles for at least a while due to other commitments. Dougweller (talk) 09:01, 20 April 2015 (UTC)

I was talking primarily about the older articles rather than new pages, yup. Articles on villages in the middle of Andhra Pradesh or northern Pakistan for instance nobody in the western world really knows about but attract local ips spewing nonsense about the famous locals and beauty of the village, a MASSIVE list of schools and businesses in cap locks and then signing the article with their own name and email address. Because few decent editors, even good Indian ones, ever see them they remain infested and get so long and bloated with haphazardly constructed factoids and POV nobody wants to clean them up.♦ Dr. Blofeld 09:41, 20 April 2015 (UTC)

The problems with India/Pakistan articles run much deeper than brevity. In fact, quite often brevity is the one thing they lack. Reams of puffery are pretty common, often copy/pasted from dodgy websites etc. I've been banging this drum since the ill-conceived WMF "push" in India some years ago but no-one seems to pay much attention and things would appear to be getting worse, as Blofeld predicts. - Sitush (talk) 09:18, 20 April 2015 (UTC)

It's going to gradually spread to other parts of Asia and Latin America and Africa too. In fact the other day I looked through some Honduran and Mexican municipalities and found some with very poorly written unsourced prose, some looked as if they'd been machine translated on the old Systransoft translator! The POV issue does seem a lot worse for India, Pakistan and Bangladesh not to mention the BLP issue as Doug identified. I've encountered a few articles on politicians in which there's been mention of corruption and scandal and unsourced! I envisage a lot of the Nigerian articles for instance going the same way. Nigerian editors are becoming more and more apparent. Sorry to be cynical. While we might attract some educated fluent editors from the developing world it's obvious that many who come online will have a poor command of English and what wikipedia really requires and will add to the current problem. The gigantic cleanup task needed and protecting articles from spam is already way beyond our control, by the 2020s it's going to be several times larger than now. At this rate in 20 years time most of our articles on the developing world will be of a shockingly poor standard as we won't be able to control what is being edited.♦ Dr. Blofeld 09:49, 20 April 2015 (UTC)

Many of these editors cannot or do not make the distinction between what is encyclopedic and what is (justifiably) important to them. Thus you are informed of where to catch rickshaws, where to get the best dosas, who are the doctors in town, and the ever-present schools available. Outreach programs should carry around a volume of any decent old-school encyclopedia and leave it with the local editors as an example of the tone and content we expect. Might not do much, but can't hurt. --NeilN ^{talk to me} 21:33, 20 April 2015 (UTC)

The people who write pages in poor English almost certainly do not have English as their first language. I think that the problem is that they might not be aware that Wikipedia is available in languages other than English. --Redrose64 (talk) 22:47, 20 April 2015 (UTC)

Oh, User:Redrose64, I guarantee that those editors know very well there are other Wikipedias in many languages. They intentionally want English Wikipedia because- that's where they want to reach potential tourists. Whether you're aiming for tourists from Australia, the UK, the USA, or even within India; it is English that will reach the most people. The Times of India for instance is the world's largest English language newspaper. We in the UK or USA may not consider their English to be up to "native speaker" status or as their first language, but chances are those people you state have "poor English" and "do not have English as their first language" have been speaking it since they were born and consider it a close second language that they use quite a lot. You're judging them by too harsh a standard, but yes a standard we have to keep strict for en:wiki.— Preceding unsigned comment added by Camelbinky (talk • contribs) 01:49, 21 April 2015 (UTC)

Judging by the level of POV in some of them and the glorification of the local imam or taxi driver I'd say a lot of them know that English wikipedia is the biggest wikipedia and want to reap praises on their home town for the world to see.♦ Dr. Blofeld 09:04, 21 April 2015 (UTC)

• meta:Talk:2015 Strategy/Community consultation/2015-03-01#Dr. Blofeld <- that is the thread that Dr. Blofeld was alluding to, i think. Some really important points there.
• Regarding poor "village articles" in India/Pakistan/other regions: Do we have for those articles something similar to Wikipedia:College and university article guidelines? Clear, specific DO-and-DONTs and recommendations for "village articles" that can be linked to in edit summaries etc. ? --Atlasowa (talk) 07:00, 21 April 2015 (UTC)

• I've not seen any such guidelines for village etc articles. However, given how little people pay attention to existing policies, even when pointed out to them, I have doubts whether such a thing would make much difference.

  There is, by the way, an element of culture clash here. For example, few of us who live in the West think that having a stable electricity supply, a bank/postal facility/major highway within 20 miles/km, decent running water etc are big deals. They are big deals in some parts of the world and it is often a source of great pride in the village that they have even one of these (or similar things). Taking electricity as one example, states in India almost go to war with each other in attempts to grab or maintain grid capacity and even then it is often rationed; same for water in some areas. Thus, while not traditionally encyclopaedic, there is an argument that having a decent grid supply or an "our pride and joy" generator might be significant. That tradition has omitted them in the past may say more about the Western encyclopaedia culture than anything else. - Sitush (talk) 07:21, 21 April 2015 (UTC)

Crazy idea #986241703... Maybe Wikipedia should not actually try to be a gazetteer. Start a separate "Wikiatlas" project that contains only highly structured "infobox-like" information about villages and similar "non-notable" places. Don't give them any room for bad grammar or to post lists of inconsequential businesses, temples, schools, etc. Actual Wikipedia articles about places then get a much higher admission standard - if it's not at least a proper Start-class it may not leave Draft-space and searches are all sent to the "Wikiatlas" page. Couple that with a hard restriction that forces anyone with less than N articles (N=3 to 5?) to create only in Draft-space. Allow only proven competent editors to create articles directly in mainspace. Yes I know that goes directly against the "anyone can edit" principle but IMHO that meme in any case needs to be consigned to Wikimedia's creation mythology - as is constantly proven on social media, the vast majority of people are definitely not capable of stringing together two simple sentences without at least three glaring errors. This is just a rough idea I'm throwing out, please don't beat me up me for not analyzing the consequences unto the seven-times-seventh generation. Roger (Dodger67) (talk) 09:00, 21 April 2015 (UTC)

What about Wikitravel? Nyth63 10:03, 21 April 2015 (UTC)

Every single human settlement is by default notable, we've been through this in many different places over the course of many, many years. We are a gazetteer and we should not be lamenting that there are a billion Indians with many towns, cities, etc that we Westerners don't know or care about. What you or anyone else or even 99% of the world cares about doesn't matter towards being included in Wikipedia. We aren't paper. Please remember that.Camelbinky (talk) 20:04, 21 April 2015 (UTC)

I think most world settlements will be notable. The sourcing for a lot of the small villages in Africa and Asia at present though might be lacking. Whatever the case, these people are going to come online and we'll see an increasing number of unsourced, poor quality new articles on villages in these areas in the next 10 years. One of the bigger problems I think though is even the towns and small cities which get more traffic and keep attracting spam. Even some of the major cities in Pakistan and Bangladesh. Somebody mentioned Chinese POV being a problem but to be honest I see very few Chinese editors editing on here. I must be missing something. Most of them are too busy editing Baidu which has over 10 million articles now. I do think we'll see more editors from China on here though over the next 10 years.♦ Dr. Blofeld 09:14, 21 April 2015 (UTC)

Now that you've brought it up, I think we should be paying much more careful attention to Baidu. Using Google translate I've looked at many of their articles and they seem to maintain a pretty high standard. Don't know if there are any Chinese speakers around who would be interested in looking into it a bit more carefully. In any case, it makes the EN Wikipedia look pretty small in comparison. Pity we can't rally the English-speakers to contribute to the same extent. It might be interesting to see what makes the Chinese so keen to participate.--Ipigott (talk) 15:15, 21 April 2015 (UTC)

I wonder if there might also be a cultural difference at work relating to the way that knowledge is transmitted outside the West. Oral transmission of knowledge is much more commonplace in non-Western cultures, particularly in places where literacy is relatively low or there are significant barriers to publishing. Information about villages might be well-known orally but might not be written down. The information in articles about such places might therefore be perfectly accurate, but would be unverifiable by Wikipedia's standards. Prioryman (talk) 19:35, 21 April 2015 (UTC)

Wikipedia explained on television

The Wikimedia Foundation can produce a half-hour television segment in English, explaining to viewers (with screenshots) the most essential things that people should know about editing Wikipedia. That segment can be broadcast separately in many regions around the globe, and it can be followed by a half-hour segment during which viewers can submit their questions (by telephone or text or e-mail), and Wikipedians can provide answers on air in real time. Interpreters can be available for viewers with low levels of fluency in English. Because of anticipated high volumes of interest in various regions, the regions can be small enough to maximize opportunities for many viewers to submit their questions. The first segment and possibly various regional versions of the second segment can be posted on WikiCommons and on YouTube.
—Wavelength (talk) 23:15, 21 April 2015 (UTC)