User:Simon Dodd/Some AFD considerations

I have been a wikipedian for nearly five years. In that time, I have formed (and reformed) many views on the jungle of wikpiedia policy and process. Some of the following observations are long-simmered views; others are snapshots of attempts to resolve difficult questions in the context of specific events. Either way, this is a draft essay that attempts to share views, insights, and opinions that I think other editors may find useful.

Heat melts SNOW

WP:SNOW allows for the short-circuiting of process when the result is a foregone conclusion. Just as WP:BURO is countered by WP:PIMP, however, SNOW closes should only be used for uncontroversial matters. user:Nathan has put it well: "Very little is ever gained by summarily shutting down a discussion or process early. It leaves participants feeling belittled and dismissed [, and] ... the pointless agita generated by this sort of peremptory action is not irrelevant, it is harmful, and it is easily avoidable."

As SNOW itself says, "[i]f an issue is 'snowballed', and somebody later raises a reasonable objection, then it probably was not a good candidate for the snowball clause." In terms, that's only useful after the fact, but ex ante guidance can be distilled from it simply by asking oneself: is it foreseeable that someone might object to an early close? To be sure, WP:NOTUNANIMITY is very clear that unanimity is not a prerequisite of consensus, and that one lone holdout can't thwart consensus, but in any discussion that is controversial, where editors have made good-faith arguments on either side of the dispute, it is eminently predictable that an early close will roil the participants on the losing end, and that they will object.

In sum, one should never attempt to use SNOW in proximity to STEAM and other sources of heat. Snowball closes should be used only for uncontensted or at least uncontroversial discussions. (Derived from [1])

AFD

When to nominate

Editors are too trigger happy with deletion nominations. I once wrote a stub article that lasted nearly three minutes before it was nominated for deletion. Over the years, I've come to strongly disapprove of nominating articles for deletion while they're still wet behind the ears. For instance, our article on Juliet Davis was barely a month old when it was nominated. Although I thought it likely that Davis wasn't notable, I argued that the nomination was premature. WP:FAILN encourages editors to use less terminal measures against an article before nominating it for deletion, and I thought these should have been tried. That is particularly appropriate for a relatively new article that is otherwise unobjectionable and whose subject has at least a faint glimmer of notability. (Cf. WP:BEEF; WP:POTENTIAL.)

A better tack, I think, is the way I have approached Albert Levitt. As with Davis, I think it very likely that Levitt is not notable, and as with Davis, the article was recently-created. Instead of nominating the article for deletion, I asked the creator whether Levitt was notable. I disagreed with the reasons he gave, so followed WP:FAILN's guidance to tag the article. I may yet nominate it for deletion - but I'm giving it time.

Why is that better? The creator thought Levitt is notable. I didn't (and don't) see it. But I could be wrong. If we let the article stay, an editor could add material supporting notability, and the encyclopædia will have grown and benefited. If we delete it, and the same material later emerges, WP:RECREATE might allow the article to be recreated, depending on the material, but time and effort by the community will have been wasted on the intervening processes, to no benefit to the encyclopædia. Strangling an article at birth at AFD is potentially counterproductive and inefficient, and generally yields no benefit to counterbalance the risk.

I am a m:Deletionist and, at least in some contexts (predominantly BLP) an m:Immediatist. But editors subscribing to those viewpoints should still be able to recognize that articles need a reasonable amount of breathing room. A new article's author should be gently pushed on the question, and the article perhaps tagged if difference of opinion is intractable. But must it be nominated? Do you suppose that we are running out of AFD, that if not nominated now the deletion process will no longer be available? I propose only a pause for breath. The article could improve. Sources supplying notability may emerge - the agglomeration of knowledge added to articles is, ex vi termini, part of the wikipedia model. Overriding considerations (WP:ATTACK, for instance) will still override. And if an article shows no sign of improvement after a few months, AFD will still be there. (Derived from [2])

Precedential effect of previous decisions to keep

When the community has decided through an AFD to keep an article, that decision has force - or at least inertia. Although no policy says it explicitly, the underlying community judgment that previous decisions to keep merit deference is glimpsed through many policies and practice. For example, WP:BEGIN tells us that before nominating an article for deletion, an editor should see "if there was a previous nomination, [and] check that your objections haven't already been dealt with." Similarly, an article is ineligible for WP:PROD if it has previously been kept through AFD, and WP:NOTAGAIN reveals the rule by making an exception (exceptio probat regulam in casibus non exceptis): "An article that was kept in a past deletion discussion may still be deleted if deletion is supported by strong reasons that were not adequately addressed in the previous deletion discussion...." None of these restrictions make sense unless they are understood to reflect the understanding mentioned above.

We saw this understanding in action in, for example, Wikipedia:Articles for deletion/Criticism of Bill O'Reilly (political commentator) (3rd nomination), where many editors argued that the previous nominations should be conclusive. Although I argued that the previous nominations of O'Reilly lacked force because they did not address (as NOTAGAIN envisions) the merits of the nomination, but rather were closed on the basis that the nomination was in bad faith, there are many deletion discussions where I have expressly deferred to previous consensus to keep despite personal misgivings. See, e.g., Wikipedia:Articles for deletion/Allie DiMeco (2nd nomination); Wikipedia:Articles for deletion/Anti-globalization and antisemitism (5th nomination). (Derived from [3])

What to do (and what not to do) at AFD

Be candid

Explain what is really on your mind as the basis for your action (paradigmatically, a delete vote at AFD). Try to expressly identify and deal with the policies and concerns you have that bear on the decision (e.g. Wikipedia:Articles for deletion/Christina Koshzow; Wikipedia:Articles for deletion/Charles Davi). You don't have to make arguments or persuade anyone, just stating your position is fine, but either way you should set out why you are doing something. And you should likewise read what other people have said on the assumption that they, too, are setting out what is really on their minds. It helps you, and it helps other users. I routinely read other people's comments at AfD and end up voting at odds with my initial instinct. Identify flaws in other users' reasoning (e.g. Wikipedia:Articles for deletion/Frank Smith (fireman) (2nd nomination); Wikipedia:Articles for deletion/Traditional marriage movement) and be responsive to flaws they identify in yours (e.g. Wikipedia:Articles for deletion/Allie DiMeco (2nd nomination)). Don't hesitate to change your mind if a user puts forward an argument you find compelling or hadn't thought of, and indicate that you've changed your mind (e.g. Wikipedia:Articles_for_deletion/Houston_Tower; Wikipedia:Articles_for_deletion/Jess Cates).

Be consistent

WP:WAX and WP:OTHERSTUFFEXISTS are not policy, and shouldn't be. They represent (at least in my own view) one of the most harmful quasi-norms on Wikipedia, the discouragement of analogical reasoning and the similar treatment of similar articles, with no countervailing benefit that I can see. Similar articles should be treated alike, and according to neutral principles, otherwise we get a kind of meta-POV problem.

Accordingly, an argument that "article X exists, so why doesn't article Y" should be evaluated on its own merits, not met with a blithe citation of WAX as a sort of wikipedian thought-terminating cliché. That isn't to say that such arguments are always persuasive. For one thing, the community may simply have never considered the compared article in an AFD (OTHERSTUFF is at its zenith when the community has not decided that the other article should be kept). For another, I see users offering bad analogical reasoning all the time here, and I suspect that's why the concept got a bad name on Wikipedia. Nevertheless, although a given (and often unadorned) "if x, why not y" argument may not be persuasive, I object to WAX's abstraction from "that's an argument that's often ill-taken" to "that's an argument that's never valid."

Note, by the way, that if you take my advice about being candid, you'll find it a lot easier to be consistent. That's not a bug, it's a feature. Derived from Wikipedia:Articles for deletion/Criticism of Bill O'Reilly (political commentator) (3rd nomination)

What the fork, let's do it

When to delete a fork

Although NPOV problems aren't by themselves a reason for deletion, the question I think we should ask in such situations is this: if the salvageable material from the fork were in the main article, would I be inclined to remove it for violating one policy or another (paradigmatically NPOV and/or WP:UNDUE? If the answer's yes, a fork basically consisting of such material with perhaps some air blown in for appearances should be deleted. Derived from Wikipedia:Articles for deletion/Criticism of Bill O'Reilly (political commentator) (3rd nomination)

The remerge fallacy

When a fork is nominated for deletion, an argument will sometimes be made that there is too much material in the fork to merge back into the main article. See, e.g., Wikipedia:Articles for deletion/Criticism of Bill O'Reilly (political commentator) (3rd nomination). That argument makes a false assumption. Although there is a conceptual difference between articles and content, the content ordinarily goes to the grave with the article. I should not be mistaken to be arguing that the content of a fork should be dumped (see WP:PRESERVE), but rather, that we should not take it as a given that it must be retained. It doesn't follow that if the fork article is deleted, the content must go back to the mothership in part or whole.

"In part" is an important qualifier, too. Even if some of the content is to be retained, articles are often written at greater length and in greater detail than necessary. The movie Titanic expends more than two hours telling this story: "the ship hits an iceberg and sinks, causing the deaths of most passengers." We should aim for a happy midpoint between those extremes, and detailed exposition is not necessary beyond the point where adding words stops adding illumination. It should also be kept in mind that a fork or subarticle can and often should contain a level of specificity that doesn't belong in (and should be deleted from if added to) the main article because of WP:DETAIL. Thus, for example, I argued in Wikipedia:Articles for deletion/Resignation of Sarah Palin that I had added material to the nominated fork that I would have deleted if it was added to the main article.

Because the material in the fork can almost invariably be compressed, arguments based on WP:SPLIT in such AFDs are usually ill-taken. SPLIT provides for a valid fork if the article has gotten too long. It might be argued, then, that this article can't be deleted because we'd have to merge at least some of the content back into that other article, and it will then be too long. The obvious problem with that argument is that because the material from the fork can almost invariably be compressed, we have no way of knowing ex ante whether the remerge will create length problems or how acute they will be. What is more, if the parent article faces size constraints, deleting the fork has utility. It forces those who want to keep the material in the fork to treat the subject within the size limits of the parent article, encouraging a leaner approach to the subject on its reincorporation into the latter. Derived from Wikipedia:Articles for deletion/Criticism of Bill O'Reilly (political commentator) (3rd nomination)

People are reading Wikipedia RIGHT NOW.

Although it isn't my philosophy, there is much to be said for m:eventualism, generally-speaking. See WP:DEADLINE, WP:NOTDONE, etc. It has no place in the context of biographies of living people, however. People are reading Wikipedia RIGHT NOW. WP:BLPSTYLE is explicit in its rejection: "While a strategy of eventualism may apply to other subject areas, badly written biographies of living persons should be stubbed or deleted," and rejection is also implicit in WP:BLP's mandate to remove contentious material (more about which in a moment). (Derived from [4]).

Actions authorized or mandated by policy

Never get involved in a land war in Asia

WP:BLP requires a "neutral, encyclopedic tone" and commands that "[c]ontentious material about living persons that is unsourced ... should be removed immediately and without waiting for discussion." The three revert rule does not apply to my removal of such material, see WP:GRAPEVINE, but it does apply to adding it. Don't fight a revert war to insert poorly-sourced material; you will lose, and risk being blocked for your trouble. Seek consensus on the talk page instead.

• Addenda 1: about "poorly sourced." WP:RS limits reliable sources to "credible published materials with a reliable publication process" whose "authors are generally regarded as trustworthy or authoritative in relation to the subject at hand." Most blogs do not meet this standard. What's more, WP:SPS bars the use of blogs as sources for many purposes (but not all - more on this anon), and WP:BLP explicitly forbids "us[ing] ... blogs ... as sources for material about a living person." A contentious claim resting purely on a source disallowed by applicable policy is inherently "poorly sourced." Accordingly, policy authorizes immediate and unilateral removal of any contentious material in a BLP that rests solely on blogs, without any 3RR penalty.

• Addenda 2: about "contentious." WP:BLP authorizes only the unilateral removal of contentious material that is unsourced or poorly sourced. But what does "contentious" mean? It isn't a synonym for "negative content"; if it was, BLP would be incoherent: its directive that "[c]ontentious material about living persons[,] ... whether the material is negative, positive, or just questionable[,] should be removed" would parse as "negative material about living persons, whether it is negative, positive, or just questionable, should be removed." "Contentious" material is best understood in its ordinary sense: material that is controversial or tends towards controversy. If the material is controversial among wikipedians (i.e. one or more wikipedians challenge or argue over it in good faith), that is a good sign that the material is controversial generally. This common-sense interpretation certainly comports with the text, and its conformance with BLP's underlying concerns is bolstered by WP:GRAPEVINE ("[r]emove any unsourced material to which a good faith editor objects"), WP:PROVEIT ("any material challenged or likely to be challenged must be attributed to a reliable, published source"), and WP:HANDLE ("Questionable material about living people in any article should be removed immediately"), all of which seem to assume the same understanding.

(Derived from Talk:Sarah_Palin/Archive_56#BLP and [5])

SPLIT and process

WP:SPLIT#Procedure mandates that "the new page should be created with an edit summary noting "split content from article name". (Do not omit this step or omit the page name.)" (Emphasis in original.) That demand should be insisted upon. SPLIT contains mandatory language; if what it purports to mandate is in fact merely advisory, it should be re-written to reflect that. Insisting on this step might seem to transgress WP:BURO, but wait. Read WP:PIMP and then consider the following.

If SPLIT's mandatory language is in fact advisory, we create a zone of administrative discretion to punish users who ignore the requirement when the administrator doesn't like the edit, relying on SPLIT, or not, when they when they like the edit, relying on BURO. Users must be able to rely on policy being enforced evenhandedly, and while I have no beef with policy being enforced as a construct of both its letter and animating purpose -- I agree with much of The Legal Process and that much of it has applicability to Wikipedia -- nothing in the underlying purpose of SPLIT undermines the text of its requirement. The procedural requirement serves the purpose, acting as a clear statement rule: a split is good, but a fork is often bad and a POV fork is always bad, yet all three look the same. How should we tell them apart? When a user follows the procedural rules of SPLIT, they demonstrate that they have thought through the implications and tension between SPLIT and WP:CFORK, and establishes a presumption that they have performed a valid split rather than an invalid fork. This is more than pedantry: process matters (that's the point of PIMP, cited above). It helps the community understand what has been done and why when the issue comes up months or years later in a setting like this one, which is a jumping-off point for these discussions. Derived from Wikipedia:Articles for deletion/Criticism of Bill O'Reilly (political commentator) (3rd nomination)

Deleting other users' comments

WP:TPO lays out a general proposition. "The basic rule is: Do not strike out or delete the comments of other editors without their permission." But WP:NOTFORUM says that editors should "stay on the task of creating an encyclopedia," so "talk pages exist for the purpose of discussing how to improve articles; they are not mere general discussion pages about the subject of the article...." Does that mean you can delete another editor's comments if you feel that they aren't directed at fixing the article?

Generally: No. There are all sorts of laws that tell people not to do something without implying a cause of action for a third party to tackle someone who breaks that law. Wikipedia isn't a legal system, but the same concept applies here. NOTFORUM tells editors what to do and not to do in their own comments and contributions. It also delineates a category of article-space material that should be removed by any other editor. It does not, however, authorize you to police other editors' talk page comments, removing those that you don't think measure up. Since NOTFORUM doesn't authorize you to remove talkpage comments violating it, and TPO has a general rule forbidding it, there is no incompatibility between those rules, and both should be followed.

With that said, TPO does make exceptions to its general prohibition, and one of them incorporates a version of NOTFORUM's rule: it allows deletion of comments (even discussions) that are irrelevant to improving the article. As WP:EXCEPTIONS and common sense make clear, however, the exception must be understood in the context of the rule. The rule is, don't delete other people's comments. When you see a comment criticizing and/or proposing to change wording in an article that you like, there is an understandable temptation to make a snap judgment that the change would make the article worse. (A fortiori if it's a passage you wrote: WP:OWN can only go so far toward suppressing the instinctive attachment of writers to their own writing!) If the change proposed would make the article worse, then it wouldn't improve the article, so the comment must be irrelevant to improving the article. So I can delete it, right? By that tortured chain of reasoning, the exception swallows the rule. That won't do. To preserve both the exception and the rule, they must be understood to protect good-faith comments seeking improvement from deletion by other editors, even if those editors think the changes proposed are stupid.

So long as the comment can reasonably be understood as be a good-faith effort to improve the article, you should not delete it - and you should assume good faith. If the discussion degenerates into a slanging match, TPO authorizes deletion of comments that are personal attacks. If the discussion wanders too far afield into a general discussion of the subject rather than one focused on the article, TPO authorizes deletion of comments. It does not allow (in fact forbids) removal of comments simply because another editor disapproves of the proposed changes (cf. WP:STEAM). (Derived from [6])

The specific controls the general

WP:GNG is a safety net, not an escape hatch. It exists, in my view (although other editors differ) to ensure that every article has an applicable notability guideline, or to resolve tension between multiple applicable notability guidelines, not to override more specific and restrictive notability guidelines that apply. For instance, GNG requires only that a topic "has received significant coverage in reliable sources that are independent of the subject"; WP:ORG, however, is more restrictive, requiring not only that an organization "has been the subject of significant coverage in secondary sources" that are "reliable, and independent of the subject," but that "[t]he depth of coverage of the subject by the source must be considered." At the other end of the scale, WP:COMPOSER is considerably more liberal than the GNG, supplying notability if a person has received "credit for writing or co-writing either lyrics or music for a notable composition" or "[h]as written a song or composition which has won (or in some cases been given a second or other place) in a major music competition not established expressly for newcomers."

Wikipedia is not a legal environment, but the maxim of statutory construction that the specific governs the general (see, e.g., Morales v. TWA, 504 U.S. 374 (1992); Gibson v. Ortiz, 387 F.3d 812 (9th Cir. 2004)) seems no less instructive - indeed, persuasive - in this context, too. The mere existence of notability guidelines other than GNG overwhelmingly makes the case that GNG does not override the more specific guidelines, because if it did, there would be no need for any notability guideline other than the GNG. The "escape hatch" conception makes surplussage of the more specific guidelines, and since that can't be right, it probably isn't. (Derived from Wikipedia:Articles for deletion/Bristol Indymedia (2nd nomination); Wikipedia:Articles for deletion/Centre for Research on Globalization (2nd nomination); Wikipedia:Articles for deletion/Jess Cates)

Gone for a song

WP:NSONG tells us that "[m]ost songs do not rise to notability for an independent article and should redirect to another relevant article," and that even if a song is notable, it should only be treated in a separate article "when there is enough verifiable material to warrant a reasonably detailed article...." Accordingly, establishing notability, particularly when notability is weak, doesn't end the inquiry. There must also be sufficient content to justify a standalone article. As I understand that requirement, in order to survive scrutiny under NSONG, a standalone article must present well-sourced information that would violate WP:DETAIL if it was included in the article for the album on which it appears. If it doesn't, the article should be deleted, or a merge and redirect performed to the appropriate parent article, even if the song is notable.

This analytic structure can also be applied in the context of WP:BIO. Similar reasoning can help untangle WP:SINGLEEVENT's concern that "[w]hen an individual is significant for their role in a single event, it may be unclear whether an article should be written about the individual, the event or both." Assuming that there are no other justifications for a standalone article (WP:SPLIT, for example), I believe that when an individual is notable only for their role in a single notable event, and there is an article about the event, we should only have a separate article about the individual when it can present well-sourced, relevant information that would violate WP:DETAIL if it was included in the article about the event. (Derived from Wikipedia:Articles for deletion/Pull Me Under and Wikipedia:Articles for deletion/Ernesto Miranda)

Understanding WP:SPS

Although blogs are troublesome as sources at Wikipedia, they are not entirely banned. WP:SPS allows that "[s]elf-published material may, in some circumstances, be acceptable when produced by an established expert on the topic of the article whose work in the relevant field has previously been published by reliable third-party publications." This is inclusive language, and given its evident purpose should a fortiori be read inclusively. WP:GAME and WP:WL are explicit that the language of a policy should give way to the policy's purpose. The concern underlying SPS is explicitly-stated: that "[a]nyone can create a website or pay to have a book published, then claim to be an expert in a certain field. For that reason," emphasis added, self-published sources such as blogs are frowned upon as sources. When the author's identity and credentials on the subject are not seriously in doubt, however, the source can be used consistently with SPS. Blog-sourced material is also acceptable if it is the product of a legitimate media outfit publishing in the form of a blog, "so long as the writers are professionals and the blog is subject to the newspaper's full editorial control." Simply put, SPS recognizes the difference between SCOTUSblog and Daily Kos.

The exception to this general rule is that WP:BLP#Reliable sources forbids use of blogs "as sources for material about a living person," period. That warning encompasses any claim about a living person, regardless of the subject of the article containing it. (Derived from Talk:United States v. Seale, Wikipedia:Articles for deletion/Bristol Indymedia (2nd nomination), and Wikipedia:Reliable_sources/Noticeboard/Archive_38#Tom_Goldstein_and_SCOTUSblog_as_sources_for_Supreme_Court_articles)

Use of WP:TLDR

The fastest way to look like a clot. "How well he's read, to reason against reading!" - Ferdinand, in Love's Labours Lost, Act 1, Sc.1.

But... Write carefully

The internet is a written medium. Write well. Read and keep at hand Strunk & White, Fowler, Politics and the English Language, the works of Bryan Garner and many more.

Concision is good, but remember that concision and verbosity are not about the word count. They are about a healthy and appropriate ratio of words to ideas. (In Glen Greenwald's case, for instance, the ratio is usually ∞:0.) Some ideas, points, or tales take longer to explain thoroughly; you should never waste your reader's time, but leaving things out sometimes means wasting even more time later because you left a gaping hole in what you told them. Keep always in mind Einstein's advice: things should be made as simple as possible - but no simpler.

- Simon Dodd { U·T·C·WP:LAW } 01:34, 13 August 2009 (UTC)