How Wikipedia Works/Chapter 10

Chapter 10: The Life Cycle of an Article

So far, we've broadly examined Wikipedia from the perspective of readers and editors. Of course, a Wikipedia reader can come and go as he or she pleases, and even the most ardent Wikipedia editor abandons his or her computer from time to time. But a Wikipedia article is always on the website, day in and day out.

So, how do things look from that article's perspective?

Let's follow Artie the Article, created by Eddie the Editor. Perhaps Artie's title is Gingerbread cottage architecture, the title used previously in Chapter 6, Good Writing and Research.

Birth of an Article

Eddie types Gingerbread cottage architecture into the search field. He discovers the article doesn't yet exist, follows the Start the Page link from the search page, and composes a few sentences. He clicks the Save Page button: Artie is born.

The moment Eddie saves his new article, it goes "live" and can be linked to and discovered through the search function. But, just as importantly, its title is immediately displayed at the top of a list called Special:Newpages. This page lists the 5,000 most-recently created articles. Gingerbread cottage architecture will slide down the list for two or three days as other editors—and possibly Wikipedia administrators on patrol—review these new articles.

After this preliminary review, many outcomes are possible.

Deletion

Wikipedia articles are created in a hostile environment, and stub articles—those short compositions of just a few sentences—are in particular peril. They are no more than tadpoles in the Wikipedia pond. New articles that do not seem appropriate for the site are often flagged for deletion as they are reviewed by other editors; this is the fate of hundreds of articles a day, many of them well meaning.

Eddie should therefore keep an eye on his article. If the content is very poor—if it contains graffiti or is un-encyclopedic, or if the topic does not seem adequately referenced for notability—a red-bar template might be added to the beginning of the text, nominating the article for deletion. (There are three types of deletion nominations, each using a different template; see Chapter 7, Cleanup, Projects, and Processes.)

Eddie should not remove a deletion template himself, but he can contest the deletion nomination. For example, he can contest a speedy deletion nomination by adding the Template:Hangon template to the article just below the deletion template and then immediately arguing his case on the article's talk page.

If a deletion tag is added to an article and is not contested, Artie's future is bleak, so Eddie needs to find out about any deletion nominations as soon as they happen. Eddie can keep track of changes to his article in a few ways:

Eddie can add the article to his watchlist by checking Watch this page (this is done by default for pages you create). Eddie's watchlist will reveal nominations for deletion and other edits to the article.

Eddie can check his user talk page. Any time an article is nominated for deletion, its creator should be informed via the creator's talk page (in this case, a message should be left on User talk:Eddie). This is not fail-proof, however, as not all editors may follow this custom.

Eddie can keep a braglist in his user space—a list of links to articles he has created. If he puts the braglist on its own page (for example, [[User:Eddie/Articles I created]]), then he can click the Related Changes link from that page for a convenient list of recent edits. This solution is even better than monitoring a watchlist for keeping an eye on just the articles you have created.

If other Wikipedia editors judge the article's content as being good enough, Artie will avoid immediate deletion. Now comes the work of improving the article.

Maintenance Tagging

As they come across articles that need attention, editors tag those articles for maintenance. Many of these yellow-bar templates might be added at the beginning of the article. For instance, if the article needs formatting work or rewriting for clarity, the {{cleanup}} tag might be added; if there are no good references, it is likely that the {{unreferenced}} tag will be placed on the article. Other yellow-bar templates may be more technical, for example, {{film-fiction}}: This film-related article may fail to make a clear distinction between fact and fiction. Particular criticisms of the writing standard may appear as orange-bar templates. For example, Template:Wikify may be added if the article could use more wikilinks. Eddie shouldn't take these templates personally—he's getting feedback on his work and now knows how the article needs to improve. It is a good idea for Eddie to do what he can to fix the article in response to any such messages.

If Artie is still a stub (in other words, just a beginning treatment, lacking something essential), editors may tag him with a {{stub}} template. But since stubs are actively sorted by category (Chapter 8, Make and Mend Wikipedia's Web), this general stub template will probably be replaced by a more specific one. For example, {{fairy-tale stub}} denotes all stubs about fairy tales, and this tag would be an appropriate one to add to Gingerbread cottage architecture. One side effect of this template being added is that Gingerbread cottage architecture will be placed in a category with similar articles that still need work, such as List of mermaid supermodels and Great Pumpkin appearances in 2008.

Editing Improvements

As soon as an article is created, other editors may set to work improving and adding to the content. Basic formatting work is often done quickly. If an article is about something in the news or an ongoing event, an editor may add a blue-bar template indicating that the article is time critical. Time-critical articles are also likely to be edited a great deal.

If the article is not about a high-profile topic (the vast majority of topics are low-profile), it might not get edited for a while. Editing might also occur in fits and starts; another editor interested in substantively working on the article may not come along for months or even years. WikiProjects generally maintain a list of new articles in their subject area, so Eddie's article may be added to one of those lists and thrive from expert attention. (Gingerbread cottage architecture might be well received at WikiProject Fairytales, for example.)

Potential Merge

A typical symbol associated with merging on Wikipedia.

Artie is not out of the woods yet. He might still be merged into an existing article, for instance, Building in folklore. One editor might feel strongly that material about a common topic has been included in many Wikipedia articles and would be better presented in a single article. Sometimes a duplicate article might not be discovered for months if it is not properly categorized and linked to other articles—did someone else create Architecture of gingerbread cottages, with similar content?

The procedural side of merging was covered in Chapter 8, Make and Mend Wikipedia's Web. If an editor proposes a merge, he or she will flag the article with a purple-bar template.

If Artie is merged, then the content created by Eddie will be included in a larger article that subsumes the gingerbread cottage architecture material. Artie will not be gone but will have become a humble redirect page. Eddie should dispute any hasty merge or redirect proposals by simply discussing them on the relevant article talk pages.

Discussion and Content Tags

A reader or editor, quite possibly someone visiting Wikipedia who is not a regular, may query what the article says. Is it true? Is it the whole truth? Is it slanted? Can a reference be provided for a specific assertion? These points will likely be added to Talk:Gingerbread cottage architecture; though in some cases, remarks might be added to User talk:Eddie—let us hope politely.

This kind of input is a further chance to improve Artie's clarity and accuracy. Comments may also take the form of orange-bar templates. For example, the template {{NPOV}} indicates that someone thinks the article fails to be neutral (does not conform to the Neutral Point of View policy). Whoever added that template should also add comments indicating his or her reason. Templates raising content issues, if not totally self-explanatory, should always be backed up by talk page comments that address the problem or slant in specific terms. Without such a detailed note, Eddie might be mystified as to what needs fixing.

If Eddie is still watching the article, he should respond to all reasonable queries rather than become annoyed. Certainly simply removing a tag requesting some sort of clarification does an article no favors, unless the tag is entirely undeserved. Having a tag on an article for a while does little actual harm, and it is normal for content to be rewritten on Wikipedia, even if the issue raised only relates to cosmetic improvements in writing style.

Bots Arrive

Icon representing wikibots.

Editors who happen to be programmers can write software for making certain types of procedural edits automatically. These programs are nicknamed bots, and Artie may be visited by a slew of them over his life. Some will make spelling corrections or small formatting changes in compliance with Manual of Style guidelines, while others analyze the content of a new page—by keywords, for example—and log it to various lists kept as project pages. (The logs can be detected in the backlinks.) Any edits made by bots will be clearly visible in the article's history, just like edits made by human editors; a bot is just another type of account. (A bot's username almost always indicates that it's a bot, not a human, for example, Sinebot, the bot that goes around signing comments on talk pages when editors fail to do so.)

If Artie hasn't been categorized, Artie's first bot edit might be the addition of an {{uncategorized}} tag.

Incoming Wikilinks

In our full review of article creation in Chapter 6, Good Writing and Research, you learned that wikilinks pointing to the new article should be added to related articles (which can themselves be found through the search function). In fact, these links should ideally be added before the new article even exists.

If Eddie is experienced at creating articles, he will consider incoming wikilinks from the outset. How many pages are displayed, as soon as Artie is created, when clicking the What Links Here link on the sidebar? Eddie should check. Being born an orphan would not be so much fun for Artie. If Eddie doesn't add links to his article, someone else might add the {{orphan}} yellow-bar template to it, placing Artie in [[Category:Orphaned articles]].

If there aren't any links to the article, there could be several explanations. For example, the chosen title for Artie the Article, Gingerbread cottage architecture, might be unconventional or spelled incorrectly. It is also possible that the concept or article title is in fact mentioned in other articles but simply has not been wikilinked. After creating the article, Eddie can still create links to Artie. Creating wikilinks will also (and subtly) draw attention to his article because the wikified pages are likely on the watchlists of editors already working in related areas. Eddie should create redirects to Artie from other possible titles, too.

An Example of Keyword Analysis

The article Carter B. Magruder, about an American general, was picked up shortly after its creation in September 2007 by a bot run by Alex Bakharev. The bot added the article to 22 logs, such as User:AlexNewArtBot/OhioLog, User:AlexNewArtBot/VirginiaLog, User:AlexNewArtBot/WWIILog, and User:AlexNewArtBot/ColdWarLog. These are simply short-term lists of new articles, kept in the User namespace and created by analyzing the article text. These logs are now routinely passed onto relevant WikiProjects, similar to the style of a Google Alert. The entry in User:AlexNewArtBot/OhioLog, a list of new articles related to Ohio, was not caused by the occurrence of Ohio in the text but by the occurrence of Cincinnati, in the phrase Society of the Cincinnati (which refers to a historical association, not the city in Ohio).

Thus bots clearly have limitations: They can suggest that articles are related to a topic area when they aren't, and they can occasionally make other mistakes a human would not make. Of course, anyone can undo edits that are not helpful and remove an article from a category or log if necessary.

Artie Is Moved

A Wikipedia move is actually a rename—something Artie might experience early in life. Particular conventions sometimes govern titles, and articles are often renamed by people familiar with those conventions. (For example, a lowercase title might be capitalized or vice versa.)

A page move, carried out by someone well meaning, might draw more attention to Artie, which might, in turn, draw more incoming links. (Artie will, of course, retain all his old wikilinks even after being renamed.) A move may also create double redirects (see Chapter 8, Make and Mend Wikipedia's Web), a technical problem that should be fixed by the article mover.

In Good Times

If all goes well, other editors will develop Artie further. Suppose Fred and Greta like what Eddie has written but think the article could be developed. Fred may standardize the formatting and add wikilinks, external links, and references, improving the article's appearance and its credibility. Greta may divide the article into sections, sorting the different aspects of the topic into some more consistent, logical order. Creating a conventional lead section that tells readers quickly whatthe content covers always helps an article. Greta will have a better idea about this once she is done with the restructuring. Perhaps, in a whimsical mood, she will even take her camera into the woods and shoot some photos of gingerbread cottage architecture.

Fred's efforts at wikifying will probably leave redlinks in Eddie's article—in other words, suggestions for more articles to write to develop Wikipedia. If these redlinks provoke Harold and Isabel, two more interested editors, to create useful new articles, Artie has really arrived in Wikipedia, and Eddie has contributed to developing the overall coverage of the topic.

In Bad Times

As editors insert additional information, the article might actually get worse, stylistically! When new facts are not integrated properly, they can upend the article's structure and muddy its writing style. (Sometimes this happens when the editors adding those facts aren't familiar with the topic or aren't fluent at editing Wikipedia.) If this continues, Artie could be destined for mere mediocrity. A strong-minded editor could step in and restore an earlier, cleaner version, do a thorough re-write, or take a red pencil to incremental changes that were not, in fact, beneficial.

When adding to articles, keep the article's overall structure in mind. Although adding a new fact onto the end of a convenient paragraph is easy, integrating that fact squarely into the article is much more helpful.

On a more positive note: Articles, on average, tend to improve over time.

Bad Times, and a True Story

Beginning on the day Artie is created, an edit war could erupt over his content, a passing vandal could deface the article, or a user could accidentally delete a chunk of vital content. These risks become more acute as more people read the article. In extreme cases, an article may be protected (or semiprotected), which halts the damage but also may prevent improvement. A gray-bar template indicates an article is protected, but full protection should only be temporary.

Another, more serious danger is that Artie may be nominated for Articles for Deletion. If the content really is worthy of an encyclopedia article, this will be a nerve-wracking time for Eddie. Even if Artie should by rights survive, the result can go the wrong way, especially if the writing is poor. That would be the end of Artie, unless a salvage mission during Deletion Review succeeds. (Deleted articles should not be re-created for six months.)

One deletion debate over a new article, which happened on September 17, 2007, has become Wikipedia legend.

During the heat of the debate, the discussion about whether a business was notable enough to have a Wikipedia page seemed typical. (Employees or others associated with a business often start these types of articles, raising questions of notability and conflict of interest.) Indeed, the article's first few hours were filled with the types of wiki perils we've discussed in this chapter. We've reprinted partof the article's edit history here; you can read what happened in reverse chronological order, just as the history would display on the site. At 11:51, less than three hours after creation, the article had been nominated for deletion, and subsequent edits were attempts to improve the article in order to influence the deletion vote.

The punchline, though, is at the bottom of the article history, in the first edit: Despite the typical arc of its debate, this was not a business trying to promote itself.

Article History:

12:13, September 17, 2007 Wikidemo (Talk | contribs) (2,206 bytes) (?Description - add material) (undo)

12:00, September 17, 2007 David Eppstein (Talk | contribs) (1,745 bytes) (?External Links - another blog entry, from Jimbo's old version) (undo)

11:54, September 17, 2007 Carcharoth (Talk | contribs) (1,610 bytes) (add three more) (undo)

11:51, September 17, 2007 ^demon (Talk | contribs) (1,529 bytes) (Nominated for deletion; see Wikipedia:Articles for deletion/Mzoli's Meats.) (undo)

11:50, September 17, 2007 Cobaltbluetony (Talk | contribs) (1,312 bytes) (dunno how the tag got back on...) (undo)

11:49, September 17, 2007 Cobaltbluetony (Talk | contribs) m (moved Mzoli's to Mzoli's Meats: full name of establishment) (undo)

11:48, September 17, 2007 EVula (Talk | contribs) (1,327 bytes) (contesting prod; I think if we give this article a bit more than a couple of hours of existence, we might have something worthwhile) (undo)

11:46, September 17, 2007 Carcharoth (Talk | contribs) (1,695 bytes) (hmm, we don't have a category on butchers, I'm not surprised) (undo)

11:46, September 17, 2007 Carcharoth (Talk | contribs) (1,717 bytes) (add categories) (undo)

11:44, September 17, 2007 ^demon (Talk | contribs) (1,645 bytes) (Proposing deletion) (undo)

11:40, September 17, 2007 Cobaltbluetony (Talk | contribs) (1,277 bytes) (notability needed according to wiki standards) (undo)

11:36, September 17, 2007 Melsaran (Talk | contribs) (1,262 bytes) ("famous" is a value judgement, and it is not really relevant anyway + doesn't add anything to the article. the fact that some sources call it famous doesn't mean that we should.) (undo)

11:34, September 17, 2007 Wikidemo (Talk | contribs) m (1,269 bytes) (restoring "famous" - source says it is; other coverage suports claim.) (undo)

11:30, September 17, 2007 Grcampbell (Talk | contribs) (1,262 bytes) (how is it famous??) (undo)

11:18, September 17, 2007 EVula (Talk | contribs) (1,269 bytes) (removing G11 tag; just because we have an article on a company doesn't mean that it is spam) (undo)

11:08, September 17, 2007 Cobaltbluetony (Talk | contribs) (1,281 bytes) (spam) (undo)

11:05, September 17, 2007 Wikidemo (Talk | contribs) m (1,269 bytes) (Undid revision 158530993 by Deb (talk) rm advertising tag - this is not written as an ad; it simply reports sourced info) (undo)

11:03, September 17, 2007 Deb (Talk | contribs) m (1,280 bytes) (tag) (undo)

11:01, September 17, 2007 Wikidemo (Talk | contribs) (1,269 bytes) (write new article; have not seen deleted version but this is new, sourced content that claims importance/notability of subject) (undo)

09:55, September 17, 2007 ^demon (Talk | contribs) deleted "Mzoli's" ? (CSD A7 (Corp): Article about a company that doesn't assert significance)

09:37, September 17, 2007 Jimbo Wales (Talk | contribs) (275 bytes) (just collecting some links as a base for writing more) (undo)

09:33, September 17, 2007 Jimbo Wales (Talk | contribs) (206 bytes) (just a stub for now, will be adding pictures and more in coming days... I need help finding reliable sources though)

That final line, at 9:33, reveals the original creator of the article: one Jimmy "Jimbo" Wales, founding father of the Wikipedia site. Jimbo's original stub of a few lines was speedily deleted after around 20 minutes on the site!

During the first few hours that Wales' article about a celebrated South African restaurant existed, it was put through all three deletion processes mentioned in Chapter 7, Cleanup, Projects, and Processes. First, the article was speedily deleted and then re-created by another author. Then it was tagged as advertising and then untagged. It was tagged as spam and nominated for a second speedy deletion. This nomination was contested; the nominator then nominated it again for proposed deletion (PROD).

This proposed deletion was again contested, so the deletion nominator called for the final deletion process at Articles for Deletion (AfD). AfD debates are intended to last five days, so this debate went on at Wikipedia:Articles for deletion/Mzoli's Meats (note that the content only exists in the page history, as this page was wiped blank and protected, an example of a courtesy blanking). The debate was closed on September 19, 2007, and the article was kept. Mzoli's Meats eventually grew into a substantial piece, with a photo and a dozen references.

This article went through the entire deletion process and survived. Its story was even featured in the LA Times. [24] Other articles are not so lucky; most pieces suspected of being covert promotion—or simply of being about non-notable topics—will face some deletion attempts, and many of these attempts will succeed. The majority of deletion debates are not as controversial as this example, however; this debate garnered extra attention both because the article was begun by Wales and because it became something of a celebrity cause within the community, discussed on the Village Pump (see Chapter 12, Community and Communication) and in one of the ongoing mailing-list discussions about how deletions are conducted.

[24] See David Samo, "Wikipedia Wars Erupt," Los Angeles Times, September 30, 2007, http://www.latimes.com/entertainment/news/la-ca-webscout30sep30,0,344107.story?coll=la-home-center

Search Engines Find the Article

Search engines often index Wikipedia articles. Here a Google search shows a number of Wikipedia articles.

As soon as an article such as Artie is created, it will be indexed and findable using Wikipedia's built-in search. General-purpose search engines will be able to find Artie within a few weeks or sometimes even sooner—within a day or two for Google. Once registered by search engines, Artie will have much greater prominence on the Web. People searching the Web generally, not specifically looking for a Wikipedia article, will begin to read the article as it shows up in their search results. This can have a good or bad effect on the article's quality. Experts in the topic may contribute to it, but random outsiders could also commit vandalism. For Artie's sake, we hope that he is now on some attentive editors' watchlists.

New Relatives

Many websites pull information from Wikipedia. Here a google knowledge panel displaying information from Wikipedia can be seen.

Over time, Artie's content will propagate outward in two ways: It will be copied verbatim, and it could be translated into other languages.

Within about a month, non-Wikipedia websites will grab Artie's content and reprint it in full. These mirrors are perfectly legal as long as they respect the GFDL license. As these copies spread across the Web, Artie's content on Wikipedia can stay more relevant by being continuously updated.

Artie's content might also be translated to provide content for Wikipedias in other languages. These new wiki articles may attract further edits, and their content may begin to diverge from the original article.

One caveat: If Artie contains mistakes, so will Artie's mirrors and translations—and even if the mistake is fixed in Artie, it will remain on the mirror sites (at least until the mirror sites refresh their content from the latest version of Wikipedia) and in translations (until someone corrects it manually).

Getting the Picture

A picture is a worth a thousand words.

Eddie might be tempted to find a relevant image somewhere on the Internet, upload it to Wikipedia, and add it to his article, but unless the image is public domain or GFDL-licensed (and most are not), this addition will lead to nothing but trouble. Non-free images are aggressively deleted from Wikipedia; they live only as long as cut flowers—a few days at most.

Instead, Eddie might take a photograph himself (or create a graphic). Or he might search for relevant images in other Wikipedia articles or on Wikimedia Commons.

Good Article

The symbol used for good articles on the English Wikipedia.

Assuming Eddie's article avoids the hazards discussed so far, Artie may be developed by skilled editors who know how to improve pages step by step. And apart from regular editing, these editors might send Artie through a number of structured improvement processes, for example, a formal peer review. If Artie is officially recognized as a good article, more possibilities open up, including induction into that rarified stratum called featured articles. The main page beckons!

Article Quality

An example of an Wikimedia article quality assessment scale.

Only about 0.1 percent of articles qualify as featured, so this parable is a little on the optimistic side, and a fairy godmother would come in handy. Some dedicated editors aim to produce featured articles from scratch, but most articles obtain this standard of quality gradually.

Other editors produce large numbers of shorter pieces. These approaches are complementary, from the point of view of the encyclopedia, and suit different editor temperaments. What really matters is that articles are in the end written collectively and thoroughly. Wikipedia has no one correct way to write an article.

Summary

The whole system for producing Wikipedia's content might seem cockeyed or random. It is certainly fallible. Content emerges from a complex but well-meaning development process, where two steps might be taken forward and then one step taken backward. But Wikipedia offers many layers of review and improvement, even if there is no single set of procedures, and ultimately Wikipedia draws readers because its content is, on balance, very useful. Indeed, Wikipedia's footprint on the World Wide Web is growing steadily.

Conclusion to Part II

Wikipedia needs all types of articles. What should you write about? Many possibilities for new articles exist, but the majority of Wikipedia's two million existing articles still need work as well. If you want to help out with articles, you can always apply basic wikification and formatting (Chapter 5, Basic Editing), work on rewriting for clarity and referencing facts (Chapter 6, Good Writing and Research), do cleanup tasks (Chapter 7, Cleanup, Projects, and Processes), sort out hypertext and category issues (Chapter 8, Make and Mend Wikipedia's Web), and perhaps even contribute images or expert syntax (Chapter 9, Images, Templates, and Special Characters).

But our central advice on writing for Wikipedia is that four things matter most to an article:

Being fairly well written and reliably sourced from the very first revision helps stave off possible deletion and provides the foundation of a good article.
Complying with the basic content policies of Neutrality, Verifiability, and No Original Research.
Fitting well with existing Wikipedia material so that incoming wikilinks exist or can be created.
Being about a topic that is also of interest to others, perhaps fitting within the scope of a WikiProject, so that others will find and develop it.

Good editors can often find parts of Wikipedia that are currently undeveloped, where new articles might be created. Just as often, many articles exist, but they are in poor shape and require a big structuring and linking effort. When you feel ready to tackle this type of work, you have mastered this part of the book. You can then consider yourself an advanced editor.