Wikipedia talk:Deletion of unreferenced BLPs

Overturning deletion

If we are going to do this - I still think we shouldn't - then it should be possible to overturn the deletion simply by providing a reference for the article (in a similar way to PROD), rather than having to go through DRV. Hut 8.5 10:07, 21 January 2010 (UTC)

I'm not totally sure about this (but I've only been thinking about it for a couple of minutes!), but what happens if the creator just removes the {{dub}} tag - is it treated like a creator who removed a SD tag from the article, in which case it is replaced and a warning issued? What happens if another editor removes the tag without adding sources - is it treated like a contested PROD - or is a warning issued?

I think these 'little' details need to be worked out, otherwise someone could just remove the tag and that'd be it! -- PhantomSteve/talk|contribs\ 10:17, 21 January 2010 (UTC)

Oh yes, those details need to be worked out. I saw it more like between an AfD tag and a Prod tag: removing it without improving the article (or without showing that the article was not eligible in the first place) should be treated as removing an AfD template: but removing it after the article has been improved should be treated as removing a prod template, i.e. perfectly acceptable, and article can not be re-dubbed (but can still be AfD'ed). Fram (talk) 10:25, 21 January 2010 (UTC)

Id the tag is removed, treat it like removal of a Prod, and allow AfD to proceed normally. Anyone abusing the process by removing the tag without adding at least one ref will quickly be found out. Collect (talk) 11:36, 21 January 2010 (UTC)

I agree completely with Fram's framing of tag removal (akin to AfD tag removal if nothing is improved, akin to PROD removal if article is sourced). Collect's suggestion doesn't really make sense to me (what are the consequences of being "found out"—does the tag go back?), and I think the only way this works is if it is absolutely verboten to remove the tag when no relevant (key word) reliable (another key word) sources are added. If anything this maybe should be spelled out more explicitly.

Incidentally I think all of the comments here are speaking to a different point than that brought up by Hut 8.5, unless I'm reading it wrong. I think the answer to that question is that DRV would not be necessary to resurrect an article. As worded the policy would prohibit recreation of the same article, and such recreations could be deleted under G4 speedy, but that only applies to "a sufficiently identical and unimproved copy," so if you added good sources you would be in the clear. Also I'm guessing it would be easy to find an admin to userfy an article deleted under this process, and then you could work to improve it and move it into article space once it was sourced. --Bigtimepeace | talk | contribs 18:01, 21 January 2010 (UTC)

Quick comment

Isn't this just supposed to be an extension of WP:BLP? WP:BLP already reads:

Material about living persons that is unsourced or poorly sourced—whether the material is negative, positive, neutral, or just questionable—should be removed immediately and without waiting for discussion.

This discussion can already be resolved with that. BLP covers removal of unsourced BLP information from ANY article, and that justifies removal of the article if it is completely unsourced. > RUL3R^>trolling_>vandalism 10:54, 21 January 2010 (UTC)

If that version of WP:BLP becomes the new policy, it may of course overrule this proposal. However, the stable version of BLP was "Contentious material about living persons that is unsourced or poorly sourced—whether the material is negative, positive, or just questionable—should be removed immediately and without waiting for discussion". Then followed a revert war and protection in the current version. There is no established consensus for the protected version of that policy yet. Fram (talk) 11:02, 21 January 2010 (UTC)

Such a reading has not been used before that I know of. Generally statements that a person exists have not been considered controversial enough for removal of the BLP in the past. Collect (talk) 11:36, 21 January 2010 (UTC)

• WP:BLP is locked because of edit-warring related to this matter and so is not reliable as a guide currently. It seems best to wait on the arbcom case before starting a novel process. Between CSD and AFD, there seem to be ample methods of dealing with any urgent examples. Colonel Warden (talk) 11:03, 21 January 2010 (UTC)

• With the current brouhaha ongoing I'm not prepared to support this. I want to mention also that I'd insist on some sort of throttle before even considering this (for example "only 1000 articles can be marked using this process at any one time") in order to prevent certain people from nominating 10,000 articles at a time. Like I said though at the beginning though, I won't support this or anything else until at least the arbcom case which brought this about is completed.
  — V = I * R (talk to Ohms law) 11:10, 21 January 2010 (UTC)

• Support Though I would likely add that holiday periods may warrant more than a strict observance of 7 days. Some editors actually do stay off-line during such periods. Collect (talk) 11:36, 21 January 2010 (UTC)

• Wait until after the current discussion at WT:PROD has run its course. A flood of different proposals and different discussions is not helpful right now. Jheald (talk) 11:45, 21 January 2010 (UTC)
• I am one of those who are quite outraged at the recent misguided deletion spree, but this proposal is reasonable. I suggest the change that someone wanting to keep an article should only have to add the reference, not necessarily remove the {{dub}} tag. Not everyone is familiar or comfortable with templates. The deleting admin should be expected to check that the article is still unsourced. --Apoc2400 (talk) 12:43, 21 January 2010 (UTC)

Yes, that was the intention. User is allowed to remove the template, but the deleting admin in any case has to check whether the article meets the requirements before deleting. The tagging for deletion may or may not be done blindly (this can be discussed), but the deletion should not be done automatically. Fram (talk) 12:59, 21 January 2010 (UTC)

• There's no harm in incubating this proposal pending the outcome of the PROD discussion and ArbCom case. The more input it gets and the longer we have to think about it, the more mature it will be if and when the time comes. I think it's a nice try but we should explicitly avoid mass taggings. Also, if we really want to fix the problem once and for all instead of in piecemeal fashion we should bifurcate this: first, go through all of the unsourced BLPs (and poorly sourced ones - which would take a sophisticated bot plus some human input) in a measured systematic fashion and either improve or delete them, and second, have a much more crisp, fast BLP deletion process for any new or newly discovered BLP articles lacking sources. I might start a parallel proposal on the first front, if none already exists. - Wikidemon (talk) 19:27, 21 January 2010 (UTC)
• Wait/Oppose While in principle support that all biographical information needs to be sourced (actually true for articles in general). However this reformulation may encourage an activism of editors (often not even familiar with the subject or sources) swarm out to remove anything from biographies that doesn't have an individual reference/footnote and would be imho a big mistake.--Kmhkmh (talk) 21:39, 21 January 2010 (UTC)
• That change to the BLP policy does not reflect consensus in any way, shape, or form, and should not be quoted. Gigs (talk) 17:23, 22 January 2010 (UTC)

Small question and suggestion

What does "dedub" stand for as in {{dedub}}? Would't {{ddub}} be more aprt, i.e. "disputed dub"? Thanks, - Wikidemon (talk) 23:50, 21 January 2010 (UTC)

Just copied from deprod. Better tags, names, ... are all welcome. Fram (talk) 14:30, 22 January 2010 (UTC)

Does not solve the fundamental BLP problem

There is suddenly a focus on "unreferenced" or "unreferenced and unwatched" BLPs. But deleting unsourced BLPs does nothing to help the BLP problem. The issue with BLPs is not a lack of references; nobody is hurt only because their accurate, fair biography is unreferenced. The risk of BLPs is with libel and inaccurate, unfair biographies. Simply adding references does not fundamentally change that risk.

A requirement for all BLPs to be sourced will only lead to people "adding sources" – which is often a euphemism for googling a random mention of the person in a newspaper and then adding that to the bottom of the article, possibly with a footnote. "Adding sources" will not fix the larger problem of sourced, biased BLPs. Unfortunately, I have seen some of the same people who are deleting unsourced BLPs encourage others to "add sources" in this way.

Moreover, even if an article is OK when initially sourced, it can still be edited at any time, at which point anyone can add libel even to a well-sourced BLP. In fact, a skilled editor can even make it look as if the libel was sourced. Many of the worst BLP articles I have seen manage to be both highly-edited and well-sourced. But even if the editor adds pure vandalism, if we already did not have the resources to watch all these BLPs, adding sources to them will not give us more resources.

Really the "unsourced BLP" issue is a result of the (real) BLP reliability issue being confused with the "cite everything" meme. But citations do not make an article reliable – editors make an article reliable. The fundamental BLP issue is that editors sometimes write libelous BLP articles, and we do not have the resources to prevent or immediately review all of them. Given that we cannot trust editors in general, why should we trust editors not to write libelous sourced BLPs?

Because of the confusion between "sourced" and "reliable", this proposal actually stands to make things slightly worse. If all the presently unsourced BLPs get sourced or deleted, some will take that as evidence that the BLP problem has been addressed, when actually it has not. It would be nice to address the real BLP problem, and I support efforts to do that through tools such as flagged revisions and increased notability standards.

But it seems to me that this proposal misses the point completely. Using "has sources" as a proxy for "reliable" is a compromise we make in ordinary articles, but for BLPs it ignores the underlying problem. It's like requiring that hemophiliacs must carry a tube of Neosporin in case they get a cut. — Carl (CBM · talk) 12:51, 21 January 2010 (UTC)

No, this proposal removes one problem, it does not have the intention of solving all other problems. I am a firm supporter of flagged revisions for all BLPs (f not even for all articles), but one could just asa easily argue that FR doesn't solve the problem of exsiting libel, incorrect claims, ... We can't solve it all in one go and with one solution, bt we can try to patch some of the major areas, and with over 50,000 articles, this is a major area. 12:59, 21 January 2010 (UTC)

My points is that unsourced BLPs on their own are no more of a problem than any other BLPs. So the proposal actually removes no problems. — Carl (CBM · talk) 13:11, 21 January 2010 (UTC)

Umm, your conclusion doesn't logically follow... The proposal removes some problems, not all of them, and has never claimed to do so. Fram (talk) 13:36, 21 January 2010 (UTC)

That's generally far too sensible (no cake for you!), but I disagree with your conclusion. Whether the proposal solves the problem depends on the extent to which articles going through it get properly sourced, rather than minimally. A "quick, here's one, can't delete" job is useless, as you say; but a proper go at sourcing everything and removing what can't be is worthwhile. Rd232 ^talk 13:37, 21 January 2010 (UTC)

Also, citing everything may not be sufficient, but it is necessary. Rd232 ^talk 13:40, 21 January 2010 (UTC)

I agree with Carl - we're tying ourselves in knots over an idea that addresses one symptom, without addressing the underlying problem. "Sourced and accurate" is our goal. "Accurate and unsourced" is better than "inaccurate and sourced". Though neither, of course, is desirable. Guettarda (talk) 14:10, 21 January 2010 (UTC)

Nice strawman. This proposal will get articles from "accurate and unsourced" to "accurate and somewhat sourced", or from "inaccurate and unsourced" to "(probably less) inaccurate and somewhat sourced". The chance that this proposal will move articles from "accurate and unsourced" to "inaccurate and sourced" is slim. So basically, this proposal will get us closer to your stated goal. If you have a separate suggestion to get rid of the "inaccurate and sourced" articles, be my guest. I fail to see how this proposal hinders you or anyone in doing that. Fram (talk) 14:19, 21 January 2010 (UTC)

My guess is that most unsourced BLP articles will go from "accurate but unsourced" to "accurate but poorly sourced". Some will become more accurate, while others will go from "inaccurate and unsourced" to "inaccurate and poorly sourced". We don't have the manpower to turn most BLP articles into "accurate" and "well sourced" ones – that is the BLP issue. Sourcing is not equivalent to accuracy, it is just a proxy that we have adopted. — Carl (CBM · talk) 14:55, 21 January 2010 (UTC)

If we don't have the manpower to turn most BLP articles into accurate and well-sourced ones, then getting rid of those that have stayed unsourced for years will let us focus a bit more on the other ones. Perhaps we also need to raise the bar of WP:BIO (and certainly subsets like WP:ATHLETE) significantly, but that's again a different (although related) discussion. Fram (talk) 15:06, 21 January 2010 (UTC)

If nobody has focused on article Y for years, deleting some other article X is unlikely to induce them to do so. — Carl (CBM · talk) 15:10, 21 January 2010 (UTC)

Honestly, if we want to claim to be an encyclopaedia, we need articles on heads of state and government, we need articles on important national-level politicians. At the same time, deleting a few tens of thousands of articles about athletes and musicians isn't going to drive contributors towards African heads of state. Guettarda (talk) 15:14, 21 January 2010 (UTC)

(ec) It's not a strawman at all, unless we have unlimited manpower (which, as Carl points out, we don't). The time spent poking around unsourced articles could just as well be spent looking at all BLPs, and working on improving their accuracy (or at least weeding out problematic content). Sure, that takes much more work. But simply pushing several thousand articled from "unsourced" to "poorly sourced" doesn't help with the accuracy thing. In fact, it takes resources away from it. If you spend your time trying to find a source - any source - for a few hundred articles, you aren't going to have much time to spend on actually improving those articles. And the time spent tagging unsourced articles will be time not spent looking for problematic content. Guettarda (talk) 15:12, 21 January 2010 (UTC)

• I think this should be put on hold until this can be discussed calmly, without mass deletions and edit warring over policy occurring, i.e. wait until the all-out war on this issue has abated! The fundamental issue is false or misleading material about living people being added and retained in any article, not just BLPs. Whether a BLP is sourced or not is immaterial to whether it is a target for such biased editing and vandalism. Removing all the unsourced BLPs will not solve the problem that a certain percentage of malicious changes are not spotted (though it will admittedly reduce the number of possible targets). Flagged revisions would help with this, as would getting more unwatched pages onto editors' watchlists. Fences&Windows 15:16, 21 January 2010 (UTC)
  • I think, beyond that, raising our notability standards some would make it easier to concentrate the effort we do have. At the moment our practices for which articles on living people are acceptable are schizophrenic. — Carl (CBM · talk) 15:22, 21 January 2010 (UTC)

+1--Kmhkmh (talk) 21:49, 21 January 2010 (UTC)

Strongly support the viewpoint that started the thread. The recent RfC showed overwhelming agreement *even among editors who are concerned about BLPs* that unsourced BLPs are not the problem, and getting rid of them will not help with the real problem. --Jorge Stolfi (talk) 21:05, 8 February 2010 (UTC)

"Consider informing"

 Done I think one contentious area in these prods is the word consider. I suggest changing this to "inform the author, unless the author is currently indef blocked". One of the contentious aspects of what has been going on is the way that authors were not being informed, and if people have culled their watchlists or don't use watchlists how are they going to learn about this other than with a note? ϢereSpielChequers 15:45, 21 January 2010 (UTC)

I have no problem with such a change. Fram (talk) 15:48, 21 January 2010 (UTC)

+1--Kmhkmh (talk) 21:44, 21 January 2010 (UTC)

What problems will the proposal solve, what will it create

I undertook a small study examining all the entries in Category:Unreferenced BLPs from November 2006, the category was mentioned in a related discussion. The overriding and nagging concern here is of course libel and defamation, i.e. unsourced negative info. I examined all entries and found NIL such problematic info. Assume now that a automated process, a dumb bot, will prod all articles, at the rate of several thousands per week. It is beyond the capacity of the contingent of the Wikipedia editors that actually wants to build an encyclopedia, as opposed to burning it, to deal with that -- it is certainly beyond my capacity, and the way I prefer to spend my time. So lets assume that all articles are deleted. What will we then have lost?

• Stuart Madnick, a renowned MIT professor with 250 books under his belt -- (update: Wikipedia:Articles for deletion/Stuart Madnick)
• Péter Medgyessy, former prime minister of Hungary -- (update: what a shocking BLP based PROD [1])
• Ampasayya Naveen, a renowned Indian writer, I found him in Gbooks in an Indian encyclopedia
• Takashi Okamura (comedian), a popular Japanese comedian, for whom I strongly suspect is an easy pass for N
• Rachid Sfar, former prime minister of Tunesia
• James Raphael, a piano player who according to the article is mentioned in New York Times and LA Times
• Benjamin Rausseo, a Venezuelan comedian, seems that is a challenger to Chavez in the upcoming presidential election, strongly indicative of a N pass
• Razi Abedi, a Pakistani literary professor, chairman of the University of the Punjab, and for that reason alone, likely to pass N*
• Luis Royo Spanish artist, Gbook hits indicative that he is notable
• Stanislau Shushkevich former head of state of Belarus
• Otto Solymosi (former?) head of National Radio Broadcast in Hungary
• Shane Sutcliffe, won Canadian Heavyweight championship 1997
• Dianne Walker, an American tap dancer, who had received praise in Boston Herald, featured in several documentaries, would be surprised if not a clear pass for N
• Yuri (Korean singer) and probably notable. -- (update: Wikipedia:Articles for deletion/Yuri (Korean singer)

These are just the most egrarious examples, there is a long tail of articles which require substantial work finding sources. What is particularly remarkable is the complete absence of problematic BLP material, so nothing will be gained in terms of flimsy, imagined and extremely poorly argued BLP concerns , but we will loose a great number of valuable articles. As a harbinger of what is coming, Viridae has already PROD'ed a number of the articles with a "unsourced BLP, fix or delete"" indicating that he is too lazy or just couldn't care less about sourcing those articles himself. I'm also concerned about the great number of non-English articles that will go in this mass fire, for which finding online sources can be quite a challenge. I'm tired of the renegade gang of Wikipedians chanting Ceterum censeo Carthaginem esse delendam at every possible occassion, and I'm concerned they eventually will succeed. Power.corrupts (talk) 16:44, 21 January 2010 (UTC)

And now Viridae has nominated Stuart Madnick, the first one on your list, for a full AfD without making any attempt himself at adding the missing sources. I think the results of the AfD so far indicate how well this attempt to use lack of sources as a deletion rationale is going to be received more broadly. —David Eppstein (talk) 17:05, 21 January 2010 (UTC)

Even worse, we're now getting prods on poorly sourced BLPs.[2] Hut 8.5 17:53, 21 January 2010 (UTC)

• They are all going to be deleted, ArbCom have sided with mass deletion of unsourced BLPs. Fences&Windows 18:02, 21 January 2010 (UTC)

Honestly, don't worry too much about that. The Arbcomm doesn't make policy. It's still up to the community to decide. If people delete important bios, improve them. And they have not condoned the disruption, even though they chose to let it slide for now. Reasonable people can work this out. Disruption, even from super-editors, can only be tolerated so far. Guettarda (talk) 18:12, 21 January 2010 (UTC)

In response to Fences, while I hate to present a fait accompli, the fact that the Arbs are validating the recent deletions makes it all the more imperative to come up with a coherent solution like this one (which I think is better than the initial proposal at WT:PROD that I made). If we have a process to go through these systematically (and a means to delete if they are not improved), it's likely that the deleting admins will stop what they are doing. If they do not, I would think the Arbs would take a different view of the matter, pointing out that a community discussion had led to a process to address the problem the deleting admins are trying to solve, albeit much more systematically and with consensus. The long and short of it is, those who opposed the suggestions at WT:PROD need to pile on board in support of some process to clean out unreferenced BLPs (I would say this one) at which point the actions of deleting admins will be clearly unacceptable and, furthermore, there will be no way that 50,000 articles will be deleted since any process like this would give time for articles to be sourced. --Bigtimepeace | talk | contribs 18:47, 21 January 2010 (UTC)

How does it mesh with the RFC? The proposals making the running there are variation on a theme compatible with this. And this is a step more developed, ready for tweaking as we see fit. Rd232 ^talk 18:54, 21 January 2010 (UTC)

Let me ask another way: Have prior problems with actual BLP violations had any relation to an unreferenced article? -- again I would like to keep "unreferenced articles" completely apart from "unsourced contentious information". Can anybody point to a list of prior BLP problem cases, could be interesting. Power.corrupts (talk) 19:24, 21 January 2010 (UTC)

Medgyessy was, indeed, a PM of Hungary -- which was evident in 10 seconds on a NYT search. PRODs of this ilk do not benefit WP one whit. Collect (talk) 21:14, 24 January 2010 (UTC)

Use of DRV

At present, the page says: "If the article has already been deleted, please follow the Deletion review process." But given that the only criterion for deletion is a lack of (adequate) sourcing, it seems like it's a fairly high threshold for recreation. After all, articles about a former prime ministers of Bosnia was deleted last night. Having to go through DRV to create an article that's unquestionably appropriate, just because you missed a PROD, seems a tad excessive. I think it should follow the standard currently used for PRODs (with a requirement that it be adequately sourced), rather than the standard for articles deleted through AFD. Guettarda (talk) 19:06, 21 January 2010 (UTC)

I thought we always permitted recreation if the article no longer failed the inclusion criteria that led to its deletion. DRV is for those cases where people dispute if it no longer fails those criteria or allege a procedural fault in the process. If the recreated BLP is sourced, the cures the fault that led to its deletion. If the re-creator wishes to re-create an unsourced BLP (for whatever reason), I don't see a problem with making them go through the higher hurdle of DRV. MBisanz ^talk 19:11, 21 January 2010 (UTC)

Why are you requiring the article to be recreated, rather than (as with a prod) allowing the old version to be undeleted on a reasonable request (possibly one that includes pointers to appropriate sources)? Requiring all the content to be written again seems to create quite a high barrier to re-entry. —David Eppstein (talk) 19:16, 21 January 2010 (UTC)

I agree that if this is to be enacted then anybody should be ablve to overturn the deletion by providing sources. This process is supposed to be modelled after PROD, after all, and this is probably going to lead to large numbers of articles being deleted when they could fairly easily be altered to avoid deletion. If we insist on DRV for every article it's going to lead to an awful lot of DRVs. Hut 8.5 19:44, 21 January 2010 (UTC)

I took recreation to include undeleting when the objection was sourcing, since that is how DRV works in similar circumstances of discovered sourcing/notability. MBisanz ^talk 19:59, 21 January 2010 (UTC)

Wikipedia:Requests_for_comment/Biographies_of_living_people

IMO, this discussion area should be closed down and people redirected to the ongoing RfC... let's consolodate the areas where this is under discussion.---Balloonman ^{NO! I'm Spartacus!} 22:26, 21 January 2010 (UTC)

I understand the point in terms of the inadvisability of splitting discussion, but still think it's good to let this run. It's a specific proposal that could be endorsed (or not) by folks participating in the RfC, and a lot of the comments there are in favor of something "prod-like" for unsourced BLPs but without it being under the umbrella of WP:PROD. Right now this seems like that option, so continuing to refine it as the RfC runs is probably worthwhile. --Bigtimepeace | talk | contribs 07:56, 22 January 2010 (UTC)

Agreed. No effort to make this into an official policy as long as the RfC or related processes are running. But if the RfC ends with something resembling support for anything sililar to this proposal, we can present this as the place to continue working on a new deletion system. If the RfC or any other main discussion concludes that unsourced BLP deletion has not enough support, then this can be marked as inactive, failed, historical, or whatever. Fram (talk) 08:11, 22 January 2010 (UTC)

Completely disagree... this proposal assumes a certain outcome and is contrary to the ongoing RfC... it is saying, screw consensus, we will continue on the path we are taking the hell with the RfC. Now if this had been ongoing for more than a few hours before the RfC, I might over look it... but the fact that they got started on the same day, leads me to question the motives here.---Balloonman ^{NO! I'm Spartacus!} 08:41, 26 January 2010 (UTC)

Yiou can question whatever you like, but you are wrong. What could be my motive in starting this six hours before the RfC? I had no idea at the time that an RfC was incoming, not that a proposal similar to this one ould be the most successful there. My motive was to provide a solution that would be acceptable to as many people as possible, without the delusion that everyone would like it. Hey, perhaps the motive behind the RfC was to avoid giving this proposal the attention it would need to ever become policy? Or perhaps some more outlandish conspiracy can be thought out? Anyway, I created this on my own, without input, feedback, discussion, chat, or other contact with anyone else, not on Wikipedia, not on IRC, not on ome superduper secret mailing list, not in real life, ... Fram (talk) 09:01, 26 January 2010 (UTC)

But this should have been put on the back burner during the RfC. I'm not questioning the motives in starting it, but in keeping it going while there was an ongoing discussion on the subject.---Balloonman ^{NO! I'm Spartacus!} 22:46, 8 February 2010 (UTC)

Speed of tagging

I haven't included this in the proposal, but I believe that some rules should be added to be effective as long as we are going through the current backlog. Assuming that this will become the new process for dealing with long-term unsourced BLPs, we need it to be reasonably fair to both sides: no 50,000 articles at once, but no 10 a day either, or we will be working at the current backlog for ten years or more. I would suggest a maximum of 100 DUB tags per day. Thoughts? Fram (talk) 13:20, 22 January 2010 (UTC)

How are we going to enforce such a limit? We already have admins who are unilaterally deleting these articles without waiting for discussion (or consensus), so what guarantee do we have that such people will actually respect the limit? Hut 8.5 13:28, 22 January 2010 (UTC)

If they see that this project has some consensus and some teeth, they will probably be more willing to follow it. The current deletions were caused by the inability of the community to tackle these articles (despite individual efforts), and the supposed need to kickstart the process. Whether it was the right thing to do or not, it at least has the benefit that everyone has to thinnk about the issue, and that many people believe it is important enough to take some ection, instead of letting these articles rest in peace. I believe that if e have a firm consensus for this proposal, and if it turns out to be effective (in tackling the backlog, no matter if the articles are deleted or better yet improved), that mass deletions of articles outside existing processes will not be done anymore.

As to how we can enforce the limit of 100 pages, I guess that it should be some kind of gentlemen's agreement, with warnings for people who repeatedly or massively overtag. Fram (talk) 13:48, 22 January 2010 (UTC)

I think we need a throttle, partly to maintain consensus, partly to give people a reasonable opportunity to fix the problem withot being swamped by 50,000 simultaneous prods, but especially to avoid this becoming a botched rushed job where editors in salvage mode quickly past in one or two references for parts of an article without checking other potentially contentious parts. As for how we run the throttle, I think the easiest to implement would be a chronological one - each day in the clear the back log period the oldest weeks of unreferenced BLPs get tagged for deletion. Yes some weeks will see many more prods than others, but the whole thing will take about 6 months and projects and editors will know which articles to prioritise in any given week or month. It would also enable a note on various projects and authors saying that "the following unsourced BLP articles are due to be prodded for deletion on the following dates unless reviewed and sourced". ϢereSpielChequers 14:21, 22 January 2010 (UTC)

Setting up some sort of a rough schedule at the outset (as WereSpielChequers seems to be suggesting) does seem to be a good idea. The current numbers people seem to be throwing out are generally in the 100 per day or 1000 per week range (which are pretty similar obviously). I would say the key thing is flexibility and starting slow at first, then ramping up a bit as time goes on. If it seems we are moving too quickly or too slowly, it should be understood at the outset that the rate could be adjusted by discussion and consensus as we move along. I don't think we'll have to worry too much about editors blatantly prodding articles over the daily or weekly limit (though probably there will be some minor and unintentional instances of that), but if anyone is doing that I'd say they should be warned and then blocked if they continue. A lot of the mechanics of this will depend on how we log these "dubbed" articles and how we actually keep track of how many articles have been tagged per day or per week.

Another thing to consider is that these questions about speed of tagging should perhaps be separate from this overall process, which could well end up being a permanent one if adopted. At first, the process would be used to facilitate the clearing of a massive backlog, and we might want to set up some kind of a separate WikiProject which defines the rules and norms for that backlog clearing. Once the backlog is actually clear we might still be using "dub" tags, but since it would hopefully only apply to new articles (since we would ideally stay on top of the problem of unreferenced BLPs going forward), questions about rate of tagging might be largely moot. --Bigtimepeace | talk | contribs 18:37, 22 January 2010 (UTC)

The process to clear the backlog may be different than the process to deal with newly-discovered unsourced articles once the backlog is cleared. It makes sense to me to flag all of the articles poste haste - not tagging them for deletion, but establishing the schedule. The BLP unsourced template does that sort of. We could add a "review date" or "review group" field to that template, or add a new one, and then use categories to manage the flow. Wikidemon (talk) 19:24, 22 January 2010 (UTC)

If people are happy with a schedule of one week of backlog per day then I would hope it would be possible to have a bot categorise the whole lot with categories in the format "category:unsourced BLPtobeproddedafterddmm2010. Salvagers would then find it easy to prioritise their article rescue work. ϢereSpielChequers 11:01, 23 January 2010 (UTC)

A concern

I think that this definitely has merit. However, I would like there to be a short note that makes it clear that even after the {{dub}} tag has been added that unreferenced and potentially libellous material should still be removed. - Tbsdy (formerly Ta bu shi da yu) ^talk 14:33, 22 January 2010 (UTC)

Oh yes, this proposal does not replace BLP. I don't know if and how this needs to be explicitly added, but any unsourced contentious content should be removed immediately, or sourced immediately. Adding a source for a birthdate and leaving the "he has murdered his three children" unsourced is not really sufficient :-) Fram (talk) 14:43, 22 January 2010 (UTC)

Evaluating sources

One thing that this proposal needs to work out is what level of sourcing is required to avoid this deletion procedure. On the high end, of course, would be multiple cites from highly reliable sources exclusively about the subject. On the low end we have no sources at all. But let's say that an article has a source to a passing mention in a local paper. It's a source, but not a good one. In my opinion, even that should be sufficient to avoid this procedure (though not a standard PROD, AfD, CSD or other deletion procedure - not sure if we need to specifically say that this procedure doesn't rule out those others.). This should be a narrow focus procedure with a clear bright line criteria. This is what I believe should be the case any time we have admins deleting articles solely on their own judgment. -208.97.245.241 (talk) 17:41, 22 January 2010 (UTC)

I agree we need to clarify this point in the proposal. We might not be able to have a bright line criteria since editors can in good faith disagree about whether a given source is "reliable", but there was a suggestion for how to deal with this at one of the discussions on WT:PROD (I think). Basically the rule should be that if there is a real dispute as to whether the article has been sourced adequately to avoid deletion under this process, then deletion should not happen and we should instead proceed to WP:AFD. I think that would work, so long as absurd claims like "but I linked to their page on IMDB!" would not constitute an actual dispute about the sourcing and could easily be dismissed. Regardless debates about the adequacy of a source will probably be one of the more contentious aspects of the process, but I think the "when in doubt, go to AfD" rule is a good baseline for handling them. --Bigtimepeace | talk | contribs 18:50, 22 January 2010 (UTC)

This is supposed to be about unsourced BLPs, extending it to poorly sourced BLPs is a logical future step - we have loads of BLPs sourced to IMDB and personal websites, but lets try and keep the focus on the unsourced BLPs for now. ϢereSpielChequers 10:54, 23 January 2010 (UTC)

Query

How many unreferenced BLPs have ever resuklted in any libel suit against WP or WMF? Collect (talk) 13:07, 23 January 2010 (UTC)

WP:BLP has very little to do with protecting the WMF from legal liability; s. 230 of the Communications Decency Act probably does that quite acceptably already, and its existence probably means that even parties who would be otherwise inclined to sue the WMF would refrain from doing so. WP:BLP is a discharge of our moral responsibility, rather than any legal one. Steve Smith (talk) 13:12, 23 January 2010 (UTC)

How many claims of libel have been made in whatever venue whatsoever that an unreferenced BLP is libelous? What percentage of unreferenced BLPs is under that category? Collect (talk) 13:55, 23 January 2010 (UTC)

How many angels can dance on the head of a pin? Steve Smith (talk) 13:56, 23 January 2010 (UTC)

IOW, none. Thanks for the reply. Meanwhile, all the noteworthy cases about false information in BLPs that have made the newspapers concern articles with references. Seems that we are in the position of the man who went to the shrink -- the shrink asked "Why do you clap your hands every five minutes?" Reply "To keep the elephants away" "There are no elephants anywhere near here!" "See! It works!" We are being asked to clap our hands, to keep the elephants away. Collect (talk) 14:00, 23 January 2010 (UTC)

No, I'm asking you to clap your hands to get the damned angels off the pinhead. While you're at that, the rest of us will focus on the BLP problem. Steve Smith (talk) 14:01, 23 January 2010 (UTC)

It's not a problem of angels. It's a necessary estimate to understand how serious is the problem we're dealing with, to know if we're using an atomic bomb to deal with a mosquito or not. --Cyclopia^talk 14:36, 23 January 2010 (UTC)

I explained above why the number of libel suits has very little to do with why we need to enforce BLP. Collect responded by asking the same question again, with some random bolding. The problem is extremely serious, whether or not there are any libel suits. Steve Smith (talk) 14:38, 23 January 2010 (UTC)

As you are a member of ArbCom making strong statements here about how serious the libel issue is, why do you not tell us something in the way of numbers? I note Scott says it is 1 or 2 percent -- while I can aver more than 10% of referenced BLPs have strong BLP issues -- surely that makes then at least five times as important to take care of? Collect (talk) 15:50, 23 January 2010 (UTC)

Let's put this another way, with extra bolding to match Collect's standards. We have established above that the problem of unsourced BLPs has nothing to do with libel: sourcing does nothing to prevent inaccurate information, most unsourced BLPs are unproblematic, and a far greater problem in that respect is unwatched BLPs. So, we're going to have to look elsewhere for an explanation of why unsourced BLPs are a serious, serious problem demanding that we solve it immediately for the sake of the children. What I would like to know is, what is the nature of the problem? If it's not issues of libel, what problems do unsourced BLPs have that they do not share with unsourced articles of any other type? Doesn't consistency demand that we immediately delete all unsourced non-BLP articles too? —David Eppstein (talk) 16:04, 23 January 2010 (UTC)

Might you indicate in what manner the argument would not fully equally apply to all unsourced articles, regardkess of category? (My bolding supra was to indicate that I wanted to be as inclusive as possible after a post which gave a reason why suits would not be filed.) Collect (talk) 16:16, 23 January 2010 (UTC)

That's exactly what I'm wondering also, yes. —David Eppstein (talk) 16:19, 23 January 2010 (UTC)

The problem of unwatched BLPs will be instantly resolved by flagged revisions, in that changes made by IPs and recently registered accounts will not be visible to the public until approved by an established Wikipedian. I don't understand why we don't have flagged revisions yet.

I agree that poorly sourced BLPs, or specific poorly sourced/unsourced statements in sourced BLPs, are more often problematic than unsourced BLPs.

There is still a difference though between an unsourced article on a living person and an unsourced article on some species of geranium or freshwater fish. The possibility of harm is inherently greater in the unsourced BLP. --JN466 19:26, 23 January 2010 (UTC)

[citation needed]. I keep hearing over and over that this is a serious serious problem and that "the possibility of harm" is huge. But it's a problem that has been ongoing for years and I have yet to see actual evidence of harm. Why the sudden rush? In what way is this a bigger problem than, say, Category:Articles that need to differentiate between fact and fiction? All I'm seeing here is an argument from repetition; why can't you be specific? —David Eppstein (talk) 20:05, 23 January 2010 (UTC)

It is my understanding that the volunteer processes outlined here:

• Wikipedia:Contact_us/Article_problem/Factual_error_(from_subject),
• Wikipedia:Volunteer response team,
• WP:BLPHELP,
• Wikipedia:Oversight

handle a considerable number of complaints on a daily basis, especially at the info-en (Quality) queue. Perhaps one of those who works on that queue can give us an indication of how many complaints there are per day, and to what extent these involve "unsourced BLPs" --JN466 21:51, 23 January 2010 (UTC)

• Collect states above that 10% of unreferenced BLPs have BLP issues (other than being unreferenced)--I suppose other than being possibly non-notable. That is not my experience, with least with the old BLPs. I have checked about 100 by now, and except for removing some fluff, have found zero that are actually harmful (there is a limitation--I do not attempt to check performers of various sorts, having insufficient background--possibly these are a real problem area) . Collect, list some please--I assume they are deleted, but I can see them. Or email me, if it's sensitive. If I am wrong, I will say so. But perhaps you mean old and new --and here i agree there is a problem, and I'd venture that 10% of all BLPs have serious problems that would fall under the BLP policy; I would also predict that the % is higher for those having sources. I'd also predict that the same percentage of all articles mentioning living people have similar problems. Does anyone care to propose a list, for people to comment on? Some actual data would help. DGG ( talk ) 02:38, 24 January 2010 (UTC)

I fear you misread my post. I said about 10% of referenced BLPs have problems. Scott stated that a figure "in the low single digits" of unreferenced BLPs have problems (presumably 1 or 2%). I make no assertions about notability or lack thereof. Thanks. Collect (talk) 03:51, 24 January 2010 (UTC)

• • I've asked arbcom for some data at RFAR, but I have no idea whether I will get a reply. I've also asked at Wikipedia_talk:Volunteer_response_team#info-en_.28Quality.29_.E2.80.93_how_many_a_day.3F (ditto). From an earlier discussion I saw at BLP talk, I understand User:Keegan might be a good person to ask for data. I have never had any interaction with Keegan though. Someone who knows him might be better placed to obtain some info.
  • Like you, I made (a smaller number of) spot checks of unreferenced BLPs. A good half of them were obscure footballers and the like from places like Iran or Thailand, just giving basic info which club they were playing for etc. Nothing causing any apparent BLP concern. --JN466 03:30, 24 January 2010 (UTC)
  • Here an article by John Derbyshire about his own Wikipedia biography, which was "well-sourced" but, as he quite convincingly points out, slanted: [3]. Quite interesting reading. --JN466 03:50, 24 January 2010 (UTC)
• I have seen quite a few of these articles in the last few days, and only two had clear BLP problems: one which claimed the subject was sacked for corruption (and was absolutely right) and one which claimed the subject had alcohol problems (which I just removed). For comparison I also fixed one case of copyright violation and another case of plagiarism. 1-2% seems about right then, but I would guess there are just as many copyright problems as BLP problems. Hut 8.5 11:24, 24 January 2010 (UTC)

• OTRS volunteers have provided some figures on the complaints volume here. --JN466 16:04, 25 January 2010 (UTC)

Sample cases

• Here are some cases:

• Mark Grebner sued the alleged editors - outcome unclear
• Fuzzy Zoeller sued the alleged editors but these case was dropped when they could not be identified
• Ron Livingston is suing the alleged editor - case still active, it seems
• Wayne Crookes sued lots of internet sites, including Wikipedia, but lost. The article in question seems to have been Green Party of Canada
• Barbara Bauer sued Wikipedia, along with other people, but lost. The article in question seems to have been Barbara Bauer Literary Agency.

It would be good to have a systematic listing of all such cases for study but, from these cases, it seems clear that:

1. Cases may arise from any type of article, not just BLPs
2. Wikipedia itself is largely protected by the Communications Decency Act
3. Plaintiffs therefore have to sue individual editors which they find difficult because they are anonymous
4. The issue of sources doesn't seem to be much of a factor. It is the derogatory nature of the statements made which matters.

My conclusion is that, if we wish to better protect people against malicious statements made here, the remedy is not to fuss over sources but, instead, to make editors more accountable by identifying themselves. I.e. no more anonymous accounts or IP edits. Colonel Warden (talk) 11:26, 24 January 2010 (UTC)

Do you mean users should edit under or declare their real names? Cassandra 73 (talk) 11:35, 24 January 2010 (UTC)

• That would be required because our licence allows other parties to copy our work and so the edit history would have to provide a reasonable audit trail. There might be some sort of third party escrow system but it seems best to keep it simple. Citizendium does this, I gather, and so I suppose that they have worked out the details. Colonel Warden (talk) 11:44, 24 January 2010 (UTC)

There are several disadvantages to use of real names, we've already had editors stalked and harassed in real life. There is also the fairness issue with some people having rare names that mean they are easily identified by google search whilst people with surnames like Smith are not uniquely identified by their name. Proper verification that every edit was made by an identified person would require an expensive verification system, and would be a major barrier to recruiting and retaining editors. There have been attempts to build an encyclopaedia writing community using that model, but I'm not aware of a successful one. ϢereSpielChequers 11:58, 24 January 2010 (UTC)

I don't dispute that there are disadvantages - I'm just saying that this is the logical remedy for the problem which so exercises people. There would be other advantages such as the elimination of the sockpuppet problem which seems more of a daily nuisance than the libel matter. Colonel Warden (talk) 12:08, 24 January 2010 (UTC)

OTRS volunteers already declare their real names to the foundation, without these names being made public. --JN466 14:20, 24 January 2010 (UTC)

I would hazard a guess that the vast majority of poor quality articles (referenced or not) about living people are in fact helpful rather than harmful to those people's careers and mental tranquility. Among other reasons are the fact that for most purposes all news is good news, and many of these articles are either direct COI, written by fans and supporters, or are simply cribbed from resumes, bios, and other promotional literature. That's not incompatible with the argument that lack of adequate sourcing may result in our publishing untrue or generally unpublicized harmful information, it's just that there's probably a lot more help than harm. When you delete an uncontroversial unsourced bio about an author, artist, entertainer, or politician, do you really think you're doing that person a favor? I think this is a lot more important for non-BLP reasons than for BLP, namely that we should be looking at some of our weakest articles and spending some time improving or deleting them rather than just creating more and more bad articles. - Wikidemon (talk) 15:59, 24 January 2010 (UTC)

Mass hysteria?

It took me a while to find this page, and I'm not certain I know the whole story, but I sense something is afoot--a mass deletion of valid and uncontentious content. Am I wrong? And if I'm right, isn't this an issue that should be going at the top of every Wikipedia page as a notification? Robert K S (talk) 09:00, 24 January 2010 (UTC)

I think that you need to qualify valid and uncontentious with either "mostly" or "almost entirely" depending on which side of the dispute you find yourself. As for notifying everyone, after the initial phase of mass prodding we are now at a discussion phase - Wikipedia:Requests for comment/Biographies of living people. If that were to decide to resume the mass deletion of circa 2% of Wikipedia's articles then I agree a notice would be useful. ϢereSpielChequers 11:40, 24 January 2010 (UTC)

Apart from the points that WSC raises, I would also query the use of "mass". Last time I looked, the most supported proposal at the rfc had fewer than 130 supporters. Some of the projects such as WP:Football that have had panicked posts about how many articles they might lose have several times that number of members. What has happenned is that the creators of and followers of Wikidrama are having their fun. However, what has been written on the pages to date should not be taken to reflect the consensus of opinion among Wikipedians. Looking at the rfc, I notice that the pattern of supports and opposes of the various statments in the last 24 hours is quite different from that in the first 24 hours the page was open. The longer we leave things, the more the general run-of-the-mill opinion of hte project will be reflected and the less that of the adherents of gesture politics, witch hunters and dramamongers will dominate.--Peter cohen (talk) 16:19, 24 January 2010 (UTC)

It does seem to follow a pattern familiar from some past deletion campaigns. - Wikidemon (talk) 16:32, 24 January 2010 (UTC)

Wikipedia mirror/clone sites

By the time articles go through speedy, speedy is contested and declined, then through AfD, very messy content can be all over the mirror/clone sites. Their crawlers may not obey HTML noindex tags when they scrape content. It would be good to be able to hide content until "cleared". Esowteric+Talk 18:33, 24 January 2010 (UTC)

Why should we care what clones/mirrors do? Why should we shape our content or policy to have an effect on any site outside of Wikipedia? Robert K S (talk) 20:46, 24 January 2010 (UTC)

Because we have a moral responsibility? Take an article that divulges inappropriate personal details of the subject, or their children, or any of the sorts of things that we regularly oversight. The subject complains, and our oversighters remove the relevant article versions from the edit history, or delete the article altogether. By that time, however, the relevant info may already have been copied by half a dozen mirrors that are run by a bot and will never remove the information, making it accessible to google for years to come. It's like closing the stable doors after the horse has bolted. To then turn around to the subject and say, "Gee, it's nothing to do with us what these mirror sites are doing," adds insult to injury.

This is the sort of thing that flagged revisions might actually help with, provided we can put unsighted drafts out of the mirrors' reach. --JN466 21:25, 24 January 2010 (UTC)

Jayen, I don't see what your remarks have to do with the Speedy/AfD process. Material in clear violation of policy (as concerns privacy, copyright, security secrecy, what-have-you) may be removed and overseen by administrators without the due process of AfD. But that is not what is being asked for by the OP, as I interpret it (from the word "messy"). Esowteric seems to be concerned about imperfect articles (e.g. the unreferenced BLPs of the present controversy) being taken up by other online resources. It seems to me we're here to build an encyclopedia, not make artificial attempts to control what other sites crawl and index; we have a moral responsibility to what is published here, but we shouldn't fallaciously reframe issues as attempts to combat an Internet genie-out-of-the-bottle control problem. Unless my interpretation is mistaken, what Esowteric proposes is directly counterposed to WP:NOTCENSORED, as what is being asked for is a "clearing board" (or "censors board") to check and verify suitability of information before making it public. Cheers, Robert K S (talk) 07:06, 27 January 2010 (UTC)

"Because we have a moral responsibility?" Also a legal one, perhaps, in the sense that I think levels of damages that might be awarded in legal actions can be affected by such secondary publication. Peter jackson (talk) 11:27, 27 January 2010 (UTC)

I don't get the issue. Anytime you post anything anywhere it gets archived and indexed somewhere. The notion that a site hosting free content has a special liability for defamation or other torts committed by its members when people reuse that content seems speculative. At most it sounds like a fairly arcane legal issue best left for the Foundation's lawyer. BLP policy already goes beyond what's required by law. If we're skirting the edge, I think it's their responsibility to tell us. Meanwhile, the notion of hiding information until "cleared" sounds a lot like flagged revisions, which a lot of people think is a good idea anyway. I too think it would be a good idea too to wait until a new BLP has been checked to make sure it's sourced before it's made available. Ironically, that possibly attaches some liability to whoever is doing the checking, if they themselves are seen as publishers rather than falling under the good Samaritan provision under the DMCA. - Wikidemon (talk) 15:30, 27 January 2010 (UTC)

I find Robert K S's attitude here quite remarkable. He asks "Why should we care what clones/mirrors do? Why should we shape our content or policy to have an effect on any site outside of Wikipedia?". Does he really think that it's alright to allow libellous material to be be spread in this way, with no attempt at all to discourage the spread? I wonder if he would still think that if he were the subject of the libel. JamesBWatson (talk) 09:58, 29 January 2010 (UTC)

If you libel someone, you are liable for that libel. You are not liable for someone who comes along and repeats the libel, no matter how ill-considered of them it may be. Others will do whatever they will do. Wikipedia is not out to police Twitter or Facebook or any other site. Wikipedia is out to create the best encyclopedia within its own confines, and others are welcome to copy or adapt content from Wikipedia so long as it is in line with the applicable license(s). I don't see the issue, legal, moral, or otherwise. (Any resource that copies from Wikipedia, and anyone that repeats what they read on Wikipedia, should be aware that Wikipedia is an encyclopedia that anyone can edit and that therefore it should not be taken to be a reliable source without further research of the facts it contains back to their individual sources.) You know when you were in first grade and you pointed at the kid across from you and yelled, "Teacher, he's coloring outside the lines!"? What did Teacher say? Teacher said, "You worry about your own work, and let your classmate worry about your classmate's work." Duly admonished then, and ditto now. Robert K S (talk) 08:20, 30 January 2010 (UTC)

Your second sentence is liable to vary between legal jurisdictions. Which ones does WMF have assets in that might be seized to enforce local court orders against it? Peter jackson (talk) 11:24, 11 February 2010 (UTC)

I agree with Robert. What mirror sites choose to do is not our concern in the least. As to the rest of the nonsense, I doubt very much that Mike Godwin has not considered this. If you aren't sure though, feel free to ask him. It isn't our responsibility to play at being lawyers. Resolute 14:45, 11 February 2010 (UTC)

True, but are the legal department watching this discussion? Have they commented here? Peter jackson (talk) 17:47, 11 February 2010 (UTC)

Notability

The article seems concerned only with BLPs that lack refs. I worry that a bio of a notable person will be deleted. For example, suppose that the Barack Obama article had no refs. Then under the proposed policy, anyone could delete it, and WP would be much poorer as a result. Of course, you may object that that article, as well as those for most notable people, does have refs. But what about an article about a notable person that was added many years ago, before it was realized how valuable refs can be? I think the policy should include an exception for some level of notability, along with guiding examples of notable and non-notable persons (I volunteer as an example of a non-notable person). David spector (talk) 15:06, 3 February 2010 (UTC)

The more notable a person, the easier it should be to source the article, and probably the easier it would be to find editors who have heard of the subject (although the latter is not needed at all to source or even create an article, but it makes it even easier to do so). And of course, claims of notability need to be sourced as well. Fram (talk) 15:34, 3 February 2010 (UTC)

And the easier it should be for the person who wants to tag the article as unreferenced or for deletion to find basic refs to preserve the article per WP:SOFIXIT.--Peter cohen (talk) 11:13, 4 February 2010 (UTC)

AFAICT, notable people were, indeed, tagged. Collect (talk) 11:30, 4 February 2010 (UTC)

Of course notable people were and are tagged. The tag is about being sourced or not, not about being notable or not. As for sofixit: tagging the article makes it clear to everyone that there is a page with a potential problem. However, sofixit only applies to the one doing the trouble of detecting the problems, apparently... I have tagged over 400 pages as unsourced yesterday: please tell me how I would have been able to source these in the same time? We all have a limited amount of time to spend on Wikipedia, and we spend it anyway we like. Sofixit doesn't mean that , you or anyone must fix a problem, but that every editor may do so. Fram (talk) 11:52, 4 February 2010 (UTC)

In other words, you don't mind if you cause other people to have to do a lot of actual work looking up and sourcing those 400 articles, but you don't want to have to spend any effort on it yourself — you'd rather just supervise. Charming. —David Eppstein (talk) 16:53, 4 February 2010 (UTC)

If I didn't tag these pages, other people still would need to do the work on sourcing them, no? I have not created any work, I have highlighted existing work so others don't have to go looking for it. I am not "supervising" anything, I haven't even put these pages on my watchlist. And it's not as if I don't do anything else outside the tagging. If you don't want to spend anytime on them, no problem, no one will blame you. How does it hurt you or anyone that these pages are tagged as unsourced? Fram (talk) 08:06, 5 February 2010 (UTC)

I have come across similar arguments before to the one David Eppstein uses here, and I have never understood them. Calling attention to a task that needs doing is a helpful and constructive action. Yes, actually performing the task would be even more helpful, but there are often good reasons why one doesn't. In any case, since we are all doing work here voluntarily I don't see how any one of us can reasonably be criticised for not doing even more. JamesBWatson (talk) 15:31, 5 February 2010 (UTC)

If the tag is only for calling attention to problem articles, as it has been in the past, then tagging problem articles with it quickly, without making improvements oneself, is useful activity and I wouldn't want to discourage it. What I was reacting to was more the recent environment in which the tag is proposed to set off some sort of short-fuse deletion process for the article. I think anyone tagging articles for deletion should follow WP:BEFORE. —David Eppstein (talk) 16:02, 5 February 2010 (UTC)

A wikiproject page for the work of referencing BLPs

I've added Wikipedia:WikiProject Unreferenced Biographies of Living Persons as a place where people interested in working on the preemptive referencing of BLPs can gather and exchange information. The project is entirely orthogonal to the proposal being discussed here, people who are interested in this may want to know about that project too. --Alvestrand (talk) 16:08, 7 February 2010 (UTC)

RFC outcome

Phase 1 of Wikipedia:Requests_for_comment/Biographies_of_living_people shows a clear consensus for a PROD-like method of dealing with the unsourced BLP backlog. It was suggested at Wikipedia:Requests for comment/Biographies of living people/Phase II that this is the best place to flesh out a detailed policy proposal. I've noted the primary points that received the broad consensus below, together with the most notable objections that will need to be addressed for this to receive btroad support. Kevin (talk) 22:33, 7 February 2010 (UTC)

Details from Phase 1

This is the detail of the proposal made by User:Jehochman.

1. Any article that satisfies the attack page criteria should be deleted on sight.
2. Biographies of living persons (BLP) articles that are unreferenced should be proposed for deletion (prod).
3. Prodding should proceed at a reasonable rate to allow interested editors the chance to add sources. The volume of proposed deletions should not be unreasonably large. Discussion can establish what is a reasonable pace.
4. After five seven days, any article so tagged may be deleted, or moved to the Wikipedia:Article incubator if it shows promise.
5. Prod notices should not be removed, nor should articles be undeleted, unless proper references are added. Anybody who engages in mass de-prodding or undeletion without adding references risks a block for disruption.
6. All editors are invited to participate in this BLP cleanup campaign.

The major objections to this were:

• The WP:PROD process should not be altered, so some other name should be used
• It is open to abuse
• The timeline is not specified
• PRODding should not happen without an attempt to source the article
• Article editors need to be notified of a pending deletion
• Quality of references to be added is unclear
• Some editors disagree with deletion as a solution altogether

Comments

I think that "Article editors need to be notified of a pending deletion" isn't so much an objection as a detail that can and should be added to the proposal. We also need to add an obligation to the prodder to check the article history, otherwise it would just be too easy for vandals to get any article deleted this way. And we need to change the new article creation process so that it is clear to all editors including newbies that if their article is about a living person they have to include their sources. ϢereSpielChequers 23:41, 7 February 2010 (UTC)

Vandals are doing this at present. Xxanthippe (talk) 00:41, 8 February 2010 (UTC).

My thoughts. Kevin (talk) 00:48, 8 February 2010 (UTC)

• Using a process other than WP:PROD is reasonable, as then there is no risk of breaking something that works well. On the other hand, altering PROD may be less bureaucratic and has the advantage that Twinkle and other similar tools are already set up for it, and can handle editor notification.
• There's not much we can do to please those who disagree with deletion altogether, or those who feel that this is open to abuse.
• The timeline might be the first thing to discuss, as it may impact other aspects of the process. I think we should set a target of somewhere between 3 and 9 months to clear the backlog.
• As for the quality of references, we need to set a bright line. I suggest that we need to see at least one reference that is not one of: self-published web site; a blog; IMDB; other user-submitted web sites. This is to prevent deletion debates where the only source is Myspace or similar.

Only X Prods per day

At the RfC, there were several calls for only a limited number of articles to be tagged per day. Does anyone know of a way for us to actually enforce this? NW (Talk) 04:20, 8 February 2010 (UTC)

No technical way that I know of exists. We'll have to go with asking editors to do the right thing. Even I can do that, despite what some may think. Kevin (talk) 04:28, 8 February 2010 (UTC)

Just a random thought: The template will likely be substituted onto the page, adding the article to Category:BLP proposed deletion by 8 February 2010. Would there possibly be a way for the template to recognize (or a bot to update the template) when then category reached a maximum number? If so, the template could then start adding articles to Category:BLP proposed deletion by 9 February 2010. That would lead to some articles being tagged for a long period of time, but at least it would establish a minimum time before things would be deleted. NW (Talk) 04:37, 8 February 2010 (UTC)

Sounds good. another way might be to tag them all at once, right now. Then wait 6 months before we start deleting them. This way every article has the same chance of being fixed. Kevin (talk) 04:41, 8 February 2010 (UTC)

No, no, no. Prodded articles get fixed when they're tagged, not so much when they expire. If we tag everything at once, then we get a push to fix them up as they are tagged which will quickly taper off without making much of a dent in the cleanup, and then a lot of deletions. This option was considered in the RfC and didn't get much support; let's not recapitulate the whole thing here. —David Eppstein (talk) 04:59, 8 February 2010 (UTC)

NW's suggestion could have the same effect if people don't stop once the limit for a day is reached. We could use a bot for the tagging which wouldn't be tempted to do just a few more. Easiest way though, is to untag any that are over the daily limit, and try to persuade people to do the right thing. Kevin (talk) 05:04, 8 February 2010 (UTC)

Probably better to somehow get a bot to do the tagging at a uniform rate rather than manually tagging and untagging so that we don't get into arguments about whether the untag means we can never reprod the same article. —David Eppstein (talk) 05:23, 8 February 2010 (UTC)

Good point. So how many per day is a reasonable number? Kevin (talk) 05:26, 8 February 2010 (UTC)

Anyone know what the current average prod rate is for comparison? --Cybercobra (talk) 05:36, 8 February 2010 (UTC)

←The past 5 days has seen between 73 and 95 PRODs per day, going by the categories. The unsourced BLP category has been dropping by about 350/day since Jan 20, for info. Kevin (talk) 05:40, 8 February 2010 (UTC)

I made the suggestion on the RFC II page that we should add a (semi?)automatic prod process that only kicked in when the backlog doesn't fall as much as we'd want it to. Normal human taggers should just tag when they feel it's justified; the "distinguished process" would only kick in if the edit process doesn't reduce the backlog fast enough for comfort. --Alvestrand (talk) 07:19, 8 February 2010 (UTC)

I think if we set a schedule, a start date and agree how many days we catch up per day then we would achieve a crude but easy to understand throttle on prods per day. So for the sake of argument if we started a prod process on the 1st March this year with five days catch up per day, we could have prodded the oldest five months of BLPs by the end of March; with this system those who choose to get involved in the prodding would know that on any particular day they could start prodding a particular 5 day stretch of the backlog without having to try and source those articles (more recently tagged unreferenced BLPs would of course still be prodable - but only under the normal prod terms). This would start the process off relatively gently and it would then accelerate greatly during the 73 days when the 30,000 unreferenced BLPs form 2009 come up at an average of over 400 a day - but I'm optimistically assuming that those projects and editors who care about these articles will have already improved the ones they care about, and therefore there will be far fewer than 30,000 when we start prodding the articles tagged in 2009. ϢereSpielChequers 22:38, 8 February 2010 (UTC)

Rather than an overall limit on prods per day I would rather see one on prods by an individual a day with an adjustment by how many articles they have removed the unreffed category from. Someone who is just making a WP:POINT by tagging a load of articles for deletion without any attempt to source should be severely restricted in what they can do. I would say no more than 5 prods per day. Someone who is steadily working through articles and providing references for as many as they are prodding shouldn't have a limit as they are acting in good faith and are actually trying to improve Wikipedia by referencing preservable notable content and only putting the harder cases forward for deletion.--Peter cohen (talk) 11:36, 9 February 2010 (UTC)

More granularity by category

In my experience most BLPs are easily identified as one of several types: sports figures, scientists, entertainers, authors, politicians, etc. They also vary by country, most importantly for our purposes, English speaking and non-English speaking, as that greatly affects the ease of finding sources. I'd like to some way to capture this information in the BLP-PROD process, allowing editors to easily assign unreferenced BLP to an appropriate category. This would allow us to prioritize and let editors who are more comfortable with one class of BLP to easily find articles to work on. We could also automatically notify various projects of BPL-PRODed articles. We might even have different deletion deadlines based on category, perhaps seeking on input from different Wikiprojects. We might even ask non-English Wikipedias for assistance in reviewing and sourcing articles about people in their purview.--agr (talk) 13:15, 8 February 2010 (UTC)

Good idea. Maurreen (talk) 20:25, 8 February 2010 (UTC)

Agree that this seems to be a good idea—would it be easy to implement? If so we should do it. I'm quite certain that the BLPs we will have the most difficulty sourcing will be biographies of people who are notable primarily or solely in non-English speaking countries. As a result there may be very few sources about them in English, but a plethora of sources in the native language or languages of their country. To delete a bio of a prominent Korean newspaper columnist while retaining one on a Canadian columnist of comparable stature would be a severe instance of system bias and we should really do our best to avoid that whenever possible (of course it will happen from time to time). Indeed the prospect of deleting bios of notable people in non-English speaking countries is my main concern as we embark on this process. If at all possible we should categorize unsourced BLPs which we are having trouble sourcing by the native language of the article subject in the hope that Wikipedians with knowledge of said language can have a go at the sourcing, or at least determine that it's insufficient for us to have an article about the person. I would also be in favor of slightly relaxing deletion deadlines for articles which need to be sourced primarily with non-English sources. --Bigtimepeace | talk | contribs 20:56, 8 February 2010 (UTC)

I agree that most such articles can be categorised that way, and indeed most already are. Our category system is the basis of the reports that have been going to various projects about "their" BLPs, and I tend to assume that the people interested in basketweavers, basketball players or bassoonists have ready access to sources in their fields of interest. I recently got a couple of reports from MZMcBride of 800 or so BLPs that only have the categories living people and year of birth, and am currently working my way through them, so far its been a right old mix though with a strong sports and entertainment bias, and about 98 to 99% I can easily add to an occupational category. I have spotted a handful of ones that definitely merited deletion or removal of unsourced negative sections, and some that I've categorised have subsequently been prodded by others - I'm assuming sometimes by people who keep an eye on particular categories and understand the subject well enough to know or care whether someone who plays for a particular baseball team is notable or not... Possibly reassuringly, at least as far as the BLP saga is concerned, I'm finding less vandalism than I would normally expect to find if I was doing some of the other things I normally do but have put on one side since the BLP kerfuffle began. I rather suspect that a very high proportion of these BLPs are lovingly put together by publicists, friends and fans which is one reason why the actual harm level of these is generally so low. When I finish my current categorisation spree I might try and get hold of BLPs with only 1 category, as I suspect there are articles in that group that no project knows of. One thing that may reassure people is that I'm finding very few completely unreferenced BLPs that haven't already been templated as such, but large numbers I've tagged as {{RefimproveBLP}} because they may only have a single source and that not always a reliable one. ϢereSpielChequers 22:22, 8 February 2010 (UTC)

I'm not sure system bias against BLPs where there are few or no English sources is bad or avoidable. We are an English encyclopedia. The long run goal might be some system of certified translations of BLPs from non-English Wikipedias, but for now I find BLPs on central Europe politicians with non-English references more worrisome than unreferenced BLPs on major league sports figures. --agr (talk) 00:03, 9 February 2010 (UTC)

I made a similar suggestion to the above one at WT:PROD, that the prod reason should roughly indicate how it should be sorted, as many prodders now do. Since there was no objection or response, I changed the guideline. WP:PRODSUM shows all the prods & reasons, eventually WP:DUBSUM should do the same.John Z (talk) 10:48, 9 February 2010 (UTC)

Come on, there are many English speakers who can do that sort of checking, and many contributors to the English wikipedia have English as their second or third language. The editors who put in those references in the first place probably are examples of that. An editor who is unable to understand a source should just skip article, and let those who can do it.
The real problem is how can we attract and keep expert editors who come here to write articles, instead of just tagging and deleting them. That is unlikely to happen as long as wikipedia praises vandals for wantonly defacing and destroying the work of its legitimate contributors. All the best, --Jorge Stolfi (talk) 17:32, 9 February 2010 (UTC)

Wikipedia does not praise vandals for wantonly defacing and destroying work. What you appear to mean is that Wikipedia accepts deletion of work in ways that you think are mistaken, but this does not mean that the editors who do so are vandals, or are acting wantonly. Making exaggerated accusations against people you disagree with is unlikely to win support for your point of view. JamesBWatson (talk) 11:00, 11 February 2010 (UTC)

Dear JamesBWatson, people have just posted statistics (here or in one of the other parallel talk pages, I lost track of them) proving that essentially all "ureferenced BLPs" in the backlog are harmless bona-fide contributions, and many of them are about definitely notable people. So there is no real justification for deleting a BLP just because it is unsourced, or even because it is non-notable. The "problem" of those BLPs is simply that some people just don't *like* them. So a lot of valid contents and a lot of other people's work got was (and may be) deleted just for that reason; and I haven't seen any apologies for that, quite the opposite. Now, what would you call me if I deleted *one* article just because *I* don't like it? (You don't need to answer: I have already been called "vandal" because I removed *one* silly editorial tag.) All the best, --Jorge Stolfi (talk) 20:58, 11 February 2010 (UTC)

It would be good if "granularity" could be added to this BLP-PROD process such that the PRODing could be integrated with the Projects. Specifically if tagged articles (for BLP-PROD) could show up on Project pages so that it would be easy for project participants to see how many articles in their subject area are in the BLB/PROD process and drill down to a list of those articles with links to them. This could be either an article status level (i.e. below Stub) or another way to count and monitor suspect BLP-PROD articles from the Project page. There would have to be some mechanism to also check on which articles were not assigned to any project and try to do that as part of the initial processing. --Mdukas (talk) 16:55, 15 February 2010 (UTC)

Here is an example using Cat Scan that might help. Maurreen (talk) 21:13, 15 February 2010 (UTC)

Objection process

The project page says: "To object to and therefore permanently prevent a BLP deletion, add one or more relevant reliable sources to the article and then remove the tag from the article. You are encouraged to also:

  1. Explain why you disagree with the deletion nomination, either in the edit summary or on the talk page.
  2. ..."

Adding the source should take care of #1. Maurreen (talk) 00:35, 9 February 2010 (UTC)

Also, the project page should spell out what to do it the BLP-PROD tag is removed without adding any source. Maurreen (talk) 00:37, 9 February 2010 (UTC)

It should, but some BLPs may have inherent notability (eg prime ministers), but not have sources easily available in English. In this case an argument can be made for not deleting despite not being able to provide a source. The result would then be incubation or AFD rather than BLP-PROD deletion. Rd232 ^talk 08:51, 9 February 2010 (UTC)

Yes, it should spell that out too. Really, someone needs to kick off a summary of the issues to be resolved. Rd232 ^talk 08:52, 9 February 2010 (UTC)

how to deal with it in a practical sense

what works best is doing real work, not talking about doing work, or planning it. At first I tried to combine both, but I decided I could do more real good by actually working on the articles. We have quite enough people giving opinions, and far from enough improving the present state of articles. Myself, I've been specializing in working on articles other people fail to source and put up for deletion as I se them on PROD or elsewhere; I find I can source half of the ones I work on, but I only work on the ones that seem likely to be notable and possibly sourceable. I'm by no means alone in this, and it's the others working on this from all different points of view whose example keeps me active. It will be more productive to just work under the current system, than figure out how to improve it, when we have no way of knowing what changes will actually be helpful. If 200 people worked each on 3 articles a day, about two months would solve what has been put forth as the immediate problem. We've been arguing about that long about what to do. By now, we could have done it. If the same number kept going at even one article/day, we could deal with the new ones as they are added. DGG ( talk ) 01:50, 14 February 2010 (UTC)

You have a fair point. But there is disagreement about what "the problem" is, such as the deletions, general lack of sources, or libel potential. Maurreen (talk) 21:16, 15 February 2010 (UTC)

Competing proposal

Please join the discussion at Wikipedia talk:Requests for comment/Biographies of living people#Next step(s). Maurreen (talk) 21:32, 4 March 2010 (UTC)