Wikipedia talk:Verifiability challenges

This has at least the following problems, many at the same locus:

1. "Material has been challenged if someone expresses a concern about the information being unverifiable, original research, or factually wrong." No, those are entirely separate concerns. Unverifiable means no published source can be found (so far), and one may not even exist. Original research is much broader than this, and most often constitutes using reliable sources that provide unconnected facts and then editorially connecting them to lead the reader to a conclusion that none of the sources cited state explicitly. Whether something is factually wrong or not is about source reliability, not lack of sources, and is an interplay between WP:RS and WP:TRUTH (and a matter for talk page exploration; we don't even have a template for this).
2. "Common ways of doing this include adding templates such as ..." – This makes no sense at all in the context. This essay is about restoring, without a source, deleted material that was removed on the accusation that is unverifiable. If the material has been tagged with {{citation needed}} (the actual name of the template, BTW) then it has not been removed, and thus this essay's scenario does not apply at all.
3. "using edit summaries that say something such as ... 'I think this is original research'" – again, OR and unverifiability are different concepts, albeit with a hazy range of overlap. If someone removes material on the grounds that it is original research, this is essentially a accusation of bad-faith or incompetent editing, so if sources had actually been provided, just allegedly misused, the burden of proof is on the OR accuser to make a case that it is OR, something that is almost always an extended talk page discussion (unless people get into an inappropriate WP:REVTALK revert war). If the material is removed with an OR accusation, but it's really because no sources were cited at all for the claim(s) in question, that is not actually an OR matter, it's an unsourced content challenge, and the deleter is citing the wrong policy.
4. The essay is not clearly distinguishing between:
   1. Unverifiability challenges (that no reliable sources exist)
   2. Unsourced content challenges (that a presumptively verifiable claim is potentially controversial enough that a source citation must be provided and has not been)
   3. Failed-verification challenges (that the cited source[s] do[es] not actually contain the claim[s] it is being cited for.
   4. Source unreliability challenges (one or more sources are provided, but are questionably reliable for any of various reasons, like WP:SPS, WP:UGC, WP:INDY, etc.).
   5. Original research challenges (the often subtle presentation or even implication of conclusions that are alleged to not be supported by RS, even if the facts used to arrive at the conclusion are all sourced).

Of these five types of verifiability challenges, the last two have nothing to do with this essay, and have a burden of proof on the deleter, because sources have in fact been provided and are just claimed to be inadequate (often correctly, a matter for determination through consensus discussion). The first three are all matters that this essay's burden-of-proof-on-the-includer interpretation of policy pertains to, but the essay only addresses the first of them, so it is misnamed and too-narrowly focused on only one of the three.

"Verification challenges" would probably be better, with two sections, for the two different burdens of proof, the dividing line between the being whether citations have been provided by an editor but not yet verified by another.

Failed verification challenges are almost on the cusp between them, but WP:AGF policy demands us to take editors at their word when they say they have obtained and checked the cited source and it does not contain the claims the article says it does. This is not even categorically a questioning of an earlier editor's faith; the most common cause of this problem is an editor between the original and the challenger having inserted content between the original claim and the citation(s) for it, making it look as if the addition is also being cited to the same source(s).

The essay could also be improved by noting that the burden of proof that applies to a challenged-and-deleted content also applies to tagging-but-preserving. E.g., one is not in a position to remove a {{citation needed}} tag (and unsourced content challenge) without providing a source, but by contrast, if someone cannot convincingly demonstrate that their perception that something is OR actually is OR, they cannot insist on retaining the OR tag on the article. The {{Dubious}} tag is (when used correctly) an unverifiability challenge, with burden of proof on an editor who would remove the tag. The opposite is true of tag claiming that citations provided are not good enough (e.g. {{unreliable source?}} and its variants).

The essay also needs to account for special scenarios where the sourcing requirements are stricter, e.g. WP:BLP and WP:MEDRS.

If these problems are solved, I think the essay might be of use, though it would arguably be better to simply clarify the wording of the relevant policies and guidelines. Whether in policy or in an essay, the clarification does need to happen, for sure. But not necessarily for the reason the essay seems to be grounded in: the idea that people restoring challenged content that is said to be unverifiable, unsourced but potentially controversial, or has failed actual attempts to verify with the cited sources, is so widespread and problematic that it must be specially addressed.

I don't see any evidence of that. However, if you follow ArbCom closely, the most common dispute formula that gets litigated there, resulting in massive amounts of drama, topic bans, long-term blocks, etc., is people removing sourced content on the claim that it contains OR or FRINGE conclusions (source misuse to advance a personal hypothesis, or use of unreliable sources to advance a hypothesis that isn't even rational), or that the sources are categorically unreliable for some reason (e.g. allegations of WP:INDY failure), then fighting half to death to stop restoration of that sourced content (or removal of tags about it) on the incorrect basis that once one of those kinds of challenges is issued, it cannot be overridden except by new sources. There is actually no support in policy for this notion, but it is widespread and demonstrably disruptive on a large scale, even if the motivation for it is usually motivated by good faith. ("Usually" should be taken seriously here. The worst disputes I've seen on WP, with the most unethical editing behavior, have been the abuse of OR policy and tags relating to it to attempt to suppress facts to suit personal or organization agendas. While the frequency with which OR claims fall into this bad-faith category is low, the disruption level generated by it is disproportionately very high.) By such burden-of-proof-reversal logic, any page could be perpetually locked in disruptive conflict by issuing challenges that every source anyone ever provides for an idea one doesn't like is unreliable (and depending on one's belief systems, this might not even be done in bad faith). Obviously, we do not actually permit that kind of negative campaigning behavior.

PS: It is pretty easy to go through the list or category of verifiability dispute templates and assign them to one of the five scenarios outlined above.

Tl;dr version: The essay is a good start, but needs work, and the input of more editors, because its policy interpretations and focus are off the mark.
 — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  20:47, 15 June 2016 (UTC)