Wikipedia:NFCC Criterion 8 debate

The problem

A discussion on the talk page WT:NFC began on March 17 2006 (the original discussion is here) and resulted, some two weeks later, in the removal of the second clause of the policy WT:NFCC#8, changing the wording from

“ Non-free content is used only if its presence would significantly increase readers' understanding of the topic, and its omission would be detrimental to that understanding. ”

to

“ Non-free content is used only if its presence would significantly increase readers' understanding of the topic. ”

Including the editor who proposed the change, a total of seven editors (including three administrators participating as editors) supported the change. Dispute has arisen as to whether this is sufficient to constitute consensus on changing policy, especially given the following sequence of events.

On April 10, administrator User:Howcheng, on discovering the change, reverted it back and started a talk page discussion. Howcheng's reversion was itself reverted, and then another administrator, User:Black Kite again reverted the change and raised the issue at the administrators' noticeboard. It is not clear whether they were acting as editors or administrators, though the venues chosen for dispute resolution suggest Howcheng was acting as an editor, and Black Kite was acting as an administrator. With an incipient edit-war brewing, the page was then protected by another administrator. Page history is here.

Two discussions are current; at WP:AN#Non-free image policy and Wikipedia_talk:Non-free_content#Criterion_8_objection where the original discussion, and the decision to alter the policy, are being debated.

This page has been set up as a centralized, and hopefully more visible, point for discussion.

Storm in a teacup?

BlackKite, there's a comment you made recently at WT:NFC that for me goes to the heart of this. I wonder if you could expand on what you wrote, and so summarise why you think these extra words add anything to the policy.

GRBerry supports, but says " If the presence of the non free content does significantly improve understanding, then it is also automatically true that omitting it would be detrimental to understanding" which, as I've said above, patently isn't the case.

Could you clarify your reasoning on this, because I think my understanding of #8 with the extra words would be much the same as GRBerry's.

Why is it that you disagree with the proposition that if the presence of the non free content does significantly improve understanding, then removing the content again would thus necessarily be detrimental to understanding that the reader would have been conveyed if the content had still been there?

As I said, that's always been my understanding of this clause, since it was added two years ago; and so why, as far as regards the meaning of the policy, AFAICS it doesn't make a lot of difference whether it's there or not. (Though for the avoidance of doubt on this point, I can see the argument which found consensus on the talk page that it might be better not to include it).

But is this not all just a storm in a teacup? Jheald (talk) 10:01, 11 April 2008 (UTC)

• Two reasons;
  • One, it's a logical fallacy, because one is not the direct negation of the other. For instance, to give a trivial analogy, if I say to you "you need to stop drinking whisky, because it gives you a hangover", you wouldn't take that to mean that you can drink all the other alcohol you can lay your hands on, because it won't give you a hangover.
  • Two: (hopefully I'll make this clear) Content can increase the understanding of an editor, even though the content is still understandable without it. Now, even if that content happens to be something that obviously violates NFCC, under the revised wording it passed #8, whilst under the original wording it often failed it. In this very common example, it will still fail another condition (probably #3a or #1), but these two strands of the policy are far more difficult to either enforce or explain to a user, because they rely far more on interpretation (for example, what does "minimal" mean? Zero images? One image?). So for those who are trying to enforce NFCC (especially with decorative fair use), the removal of the second clause robs us of a vital policy which we can point editors towards when trying (nicely) to explain why their images have been removed or reduced. (And that's why IMO it's not a storm in a teacup). Black Kite 10:52, 11 April 2008 (UTC)

"Minimal" is inherited from US law. It means "no more than required to achieve the identified purpose".

Content can increase the understanding of an editor, even though the content is still understandable without it. No. The test for #8 has never been whether the article is not understandable without it. That view may have been canvassed when the clause was first added two years ago, but if you look back at the discussion then, it was only ever held by a few, not the majority or a consensus. Rather, the test is whether there is an additional understanding a reader would gain by including the image, that would be detrimented by its omission.

Unlike your whisky example, I find it very difficult to see circumstances where you would allow that the n.f.c. significantly contributed to understanding, where removal of the n.f.c. would not then be detrimental to that improved understanding. But perhaps, instead of just citing material that "obviously violates NFCC", you have some more specific nfc in mind that would satisfy the first part of the clause but not the second? Jheald (talk) 11:27, 11 April 2008 (UTC)

First of all, yes, I realise the concept of "minimal", but of course anyone unfamiliar with US law (i.e. most people) wouldn't understand exactly what it means at all. For example, I had an editor recently - quite an experienced one - who swore that it meant "5 images or less". Secondly, I freely admit that there are few cases where the first and second cases don't overlap - but if you read above, that's not what I said - the important reason for the existence of the second clause, as far as non-free image editors are concerned, is that it provides a much simpler way to explain to editors why images are non-compliant. It is very easy to waste a lot of time in a circular conversation with someone who is claiming that their images of minor characters from a movie add to their understanding of it, but it is very difficult for editors, in the same case, to argue that their understanding would be diminished if said images were removed. This is the crux of the matter - non-free image editors have a hard time enough of it at the moment - many don't last very long doing this work - and this change to #8 makes it much more difficult for them to make articles compliant with the Foundation's policies. In the end, we have a dilemma - either we give every assistance to editors and admins who are trying to work with NFCC, and we make NFCC easier for editors to understand, or we give up and allow unlimited free use and take the word "Free" out of the top left hand corner of every Wikipedia page. That is the reality of the situation; there is little centre ground. Black Kite 16:52, 11 April 2008 (UTC)

"The test for #8 has never been whether the article is not understandable without it" -- correct. The test for #8 is whether the article is less well-understood (the former is a complete negation, whereas the latter is a difference in degree). Usually #8 is applied not to the article as a whole, but to the section that the non-free content contributes to. The common example I use is magazine covers. Frequently, we have "So-and-so was named TIME's Man of the Year for 19xx" accompanied by the image of the cover, and arguments for inclusion of that image have been along the lines of, "This shows how so-and-so was perceived by the general public, and the style of dress of the time," etc, when none of that is even mentioned in the article. You can understand Billy Ripken if you don't see his baseball card, but it just makes so much more sense with it. Demi Moore is perfectly understandable without the magazine covers, but their impact gives a sense that simply cannot be conveyed in words. howcheng {chat} 17:26, 11 April 2008 (UTC)

The Demi Moore example is out of date. That article now has no pictures. See instead More Demi Moore, Leibovitz v. Paramount Pictures Corp., and Demi's Birthday Suit. Interesting how that example has evolved as the articles got written. Carcharoth (talk) 18:03, 11 April 2008 (UTC)

Duh, I think I knew that also. howcheng {chat} 18:21, 11 April 2008 (UTC)

Role of administrators

Moved to Wikipedia talk:NFCC Criterion 8 debate#Role of administrators

Criteria for assessing consensus

Moved to Wikipedia talk:NFCC Criterion 8 debate#Criteria for assessing consensus

WP:SILENCE

Moved to Wikipedia talk:NFCC Criterion 8 debate#WP:SILENCE

Summary:

• Situation unfolded exactly as predicted by WP:SILENCE. No one should be surprised. People should Read The Fine Documentation and prevent future repeats.
• Too much talk, not enough editing.
• Objectives are to unprotect page and restore normal editing, so that we can work towards consensus.

criteria for both addition and removal

I prefer the previous language because it gives criteria for both addition and removal of images. If I want to add a nonfree image, I ask myself whether doing so presents a significant benefit to the reader. Only if the addition is significantly beneficial should the image be added. If I want to remove an image, I ask myself whether doing so is detrimental to the reader. If removing the image is not detrimental, then the image should be removed, since doing so serves the second goal of reducing non-free content.

This tension in the phrasing, with criteria for both adding and removing images, makes it more difficult for editors to stretch #8 too far in either direction. If it focused only on adding images, it would be easier to stretch in favor of their addition. If it focused only on removal of images, it would be easier to stretch in favor of their removal. — Carl (CBM · talk) 11:46, 11 April 2008 (UTC)

How about making this explicit in the wording? Split NFCC#8 into two parts: how to apply the criterion to including an image, and how to apply it to removing an image. This could be applied to all the criteria if that would help. eg. NFCC#1 to remove an image under this criterion, explain clearly what the free alternative is and how it is equivalent to the image being removed (people often forget this last bit). To add an image, explain clearly how no free equivalent can be created. For NFCC#2, state clearly (with evidence) what the original market role was, and explain what commercial opportunities are being infringed. And so on. Increase the quality of debate on either side, rather than having jaded editors and admins saying "blatant" and "obvious" and failing to explain themselves. Carcharoth (talk) 12:02, 11 April 2008 (UTC)

For #1 at least, the burden has historically been on those in favor of including an image, who have to complete a nonfree use rationale.

The deeper issue with #8 is that there is no clearly delineated level of significance that is enough to justify use of an image. The image needs to do more than just be related to the subject of the article, but doesn't have to literally be the subject of the article. So the two-pronged test helps clarify things. An image being added should be significant to the point that its subsequent removal would be detrimental to the reader. Similarly, when we consider removing an image, the degree to which the image is significant helps us decide whether its removal is detrimental. I think it's a nice balance of interest. — Carl (CBM · talk) 12:13, 11 April 2008 (UTC)

This hits one of the nails on the head. TONY (talk) 05:50, 13 April 2008 (UTC)

My thoughts on criterion 8

I much prefer version A (one sentence). The reason is that the NFCC is supposed to be an objective policy, but criterion 8 is really subjective (#1, #3a, and #5 require a bit of subjectivity but not on the scale of #8). The second line opens up a whole bunch of problems, for instance. I'm taking one example: Talk:Partners in Crime (Doctor Who)#Adipose - pics available was the second of two discussions regarding a fair use image. Looking at the discussion and at the article (for comparison, a version with no fair use can be seen here, you can say it easily passes the first line of "significance":

In this instance, I am inclined to agree with inclusion (and you guys know I'm otherwise among the most critical about episode images). If the introduction of computer-animated cute little blobs is such a notable new feature in this episode, let us see them. It really makes the coverage more informative. - Future Perfect at Sunrise

Black Kite and Serpahimblade, two other editors who are well known for being very critical about fair use, (the former is instrumental in starting this dispute) agreed that removing the picture would be inappropriate. The problem is, the wording of the latter still facilitated an edit war about the image. It's just one example of how these disputes devolve into "yes it is; no it isn't; yes it is; etc" argument. That sort of wording doesn't belong in an objective policy. Sceptre ^(talk) 12:06, 11 April 2008 (UTC)`

It is very true that we have no functional system for resolving disagreements about nonfree images. On the other hand, I don't think it will ever be possible to make a particularly objective policy about nonfree image use that even resembles our practice, which is not objective. — Carl (CBM · talk) 12:17, 11 April 2008 (UTC)

If 8 is going to cause so much trouble (the "subjectivety"), it needs greater detail in the wording. I know it's hard, but there'll be no end of trouble until 8 is written out properly. The way to start is to get the people who do the policing to describe borderline cases they've encountered, and prompt feedback from people here on where and how to establish the boundary for each case, and how to express it in words. Might be messy, but it's a start. This is often how legislation is developed by policy-makers. TONY (talk) 12:22, 11 April 2008 (UTC)

Tony, there is no end of trouble even with the extra clause of #8 included - what is it going to be like for NFCC patrollers with it removed? For an example, see Talk:List_of_gangs_in_the_Grand_Theft_Auto_series#Images and all the sections below it. This was an article which had very obvious fair-use issues - all the images were pretty much identical screenshots of gangs from the video game. It failed #1, it failed #3a, and it failed #8 in both its incarnations. Yet look at the discussion -"Basically we came to a consensus and included the 6 images we felt best represented the group as a whole". "I think your reasoning is weak, your editing sloppy, and your authoritative stance to be highly presumptuous". "Regardless of what's policy ... what gives you the right to single-handedly administer that policy?" et cetera, et cetera. And that's one of the more civil discussions :) Do you see now the reason why NFCC has to be watertight, or there's no point having it at all? Black Kite 17:08, 11 April 2008 (UTC)

And again, a list of numbers with no reasoning. For NFCC#1, if an image fails it, you should show this by suggesting what free replacement is possible. What free images are possible for List of gangs in the Grand Theft Auto series (leaving aside for a moment the points of whether the article should even exist or whether any image is needed at all). For 3a: "Multiple items are not used if one will suffice; one is used only if necessary." - for this you need to demonstrate that one image will suffice. If your argument is that no images are needed at all, you should not be referring to 3a at all. NFCC-3a is for cases where one image is useful, but people are trying to use more than one image. Strictly, the "one is used only if necessary" bit is really NFCC#8 in another form. Finally, we come to NFCC#8, and the sloppy wording means that precise arguments are no longer possible. That is why it needs to be rewritten. Carcharoth (talk) 17:16, 11 April 2008 (UTC)

No, my point here was really to show the types of arguments that NFCC patrollers needs to answer on a regular basis. In this particular case the images clearly failed all three of those clauses (#1 - description could've been replaced with text: #3a - zero or at the very most one image was sufficient: #8 - a whole slew of screenshots showing random groups of video game minor characters standing around on street corners could never increase understanding) Black Kite 17:37, 11 April 2008 (UTC)

Thank you. That is a much clearer answer. I agree, mainly because I had temporarily forgotten that written descriptions can replace decorative text. Still, montage images are an interesting edge case. See List of characters in The Simpsons. Carcharoth (talk) 18:05, 11 April 2008 (UTC)

Indeed; another good example which I was involved with here. Black Kite 18:24, 11 April 2008 (UTC)

We're not writing legislation. But I agree it might be useful to see folks documenting what decisions they've been making and why. That's always helpful. :-). Just make a link off of point 8 pointing to [[/additional documentation]] for now. If everyone agrees not to go reverting willy-nilly, can we go get an unprotect and start on that? :) --Kim Bruning (talk) 13:11, 11 April 2008 (UTC)

I'm a technodummy about links off things ... how does one do that? It sounds like a good idea. I want to make the point that if we can come up with the appropriate detail, we'll have solved THE central problem for this policy. TONY (talk) 13:34, 11 April 2008 (UTC)

Eh? You link using the pretty square brackets? :-P Hmph, I'm not entirely convinced making a subpage will work, but if it doesn't help at least it won't hurt... too much. O:-) Can we unprotect?--Kim Bruning (talk) 13:37, 11 April 2008 (UTC)

I think you need to ask the people who were edit warring. Specifically, User:DCGeist and User:Black Kite. The protecting admin, User:Kwsn, also left the following notes when changing the protection level from a week to indefinite, so probably best to ask him as well: "Protected Wikipedia:Non-free content criteria: Stop the stupid bickering already, take it to talk if you want to make changes" and "template to make it painfully obvious" and "Changing to indef on second thought". I've excluded User:Howcheng from all this, as he made one revert and went straight to the talk page. User:Tony1 obviously made a change that he thought had consensus, and, like Howcheng, didn't edit war after he was reverted. Everyone else was only discussing, so asking them if they will calm down is a bit pointless. So, as a neutral party, Kim, could you negotiate with DCGeist, Blank Kite and Kwsn to get the page unprotected? Carcharoth (talk) 14:13, 11 April 2008 (UTC)

I also object to tony's new wording, NFCC#8 was ment to be a double edged sword, and be a useful tool for both inclusion and exclusion of non-free material. One that the image is relivant and improves the understanding of the subject, and Two, that the improvement is significant. that is useful because every image improves the understanding of the reader, most provide very very little improvement, but an improvement none the less, and per the new wording removing those will be very difficult. β^{_{command 2}} 14:39, 11 April 2008 (UTC)

Overall I support the more restrictive wording, for two reasons: (1) it is a better balance between the different interests at issue, (2) at the moment our action agenda should be about removing the many excess non-free images, and this language is more supportive of that. Christopher Parham (talk) 21:47, 11 April 2008 (UTC)

I'm one who thinks the new wording arrived at in the March debate (which is apparently the old wording) represents a significant improvement. I understand the arguments for how that strangely conceived second clause helps some of our administrators rid Wikipedia of some of the more ridiculous and unnecessary images that are uploaded here. At the same time, it clearly causes problems for the many editors here who are devoted to doing quality work on encyclopedia articles that aren't about ...um...cartoons and games. I don't know about anyone else, but I think we should be doing everything we can to support our best work. That second clause should remain scrapped. In its place, as some have suggested more or less plainly above, we should add some more detail on what qualifies as "significance". That should both support our quality editors much better while still giving our enforcement administrators viable tools to work with. Two main areas of "significance" that shouldn't be impossible to define occur to me: (1) Where the image illustrates the singular topic of the article (album cover for article on album; photograph for article on person; exemplary screenshot for article on TV episode); (2) Where the image illustrates a person, object, or event important to an appreciation of a broader article topic—the significance of that person/object/event supported by sourced commentary. DocKino (talk) 00:26, 12 April 2008 (UTC)

Yet another debate on this

These debates on fair use never, ever, ever end. They never will.

I've written about this before, but there's basically two camps:

• Fair use inclusionists who wish to use fair use liberally, and do so whenever it would enhance an article to have an image. Thus we end up with the removal of a phrase in criterion 8.
• Fair use deletionists/minimalists who push to reduce fair use as much as possible. Thus we end up with the re-inclusion of the phrase into criterion 8.

Even though the Foundation has taken a strong stance that fair use should be done minimally, within narrow limits (see Foundation:Resolution:Licensing policy) and that policy can not be eroded, the inclusionists insist on liberal inclusion. They are regularly attacked for their stance. Similarly, the deletionists/minimalists are also attacked. It never ends.

So, we end up with things like List of Kanon characters which at 60kb long has 22 fair use images, or one for ever 3kb of actual text. But if you try to reduce the fair use dependency in articles like that and you'll get burned at the stake. We end up with never ending debates about the policy. We end up with revert wars over the content of the policy. Why? Because there's no middle ground.

Any consensus which concludes that fair use should be more liberally allowed is overruled by Foundation policy. The "consensus" to reduce the limitations on fair use via criterion 8 is meaningless in the face of Foundation policy. --Hammersoft (talk) 14:41, 11 April 2008 (UTC)

FYI The foundation document was actually intended as a summary of the then-current consensus. :-P --Kim Bruning (talk) 14:55, 11 April 2008 (UTC)

• That List of Kanon characters is yet another popular culture article with ludicrous amounts of fair use which clearly violate NFCC completely. I've tagged it; after a period of notice the violating images (in this case, all of them unless the article can be recast to try to fit parts of #1) will need to go. Black Kite 16:58, 11 April 2008 (UTC)

• Are you nuts? How can a person possibly understand which character is which in that series without having a picture of each character? What are we supposed to do, written extensive text descriptions how each character looks??? add smiley faces as appropriate until it is understood the preceding comments are intended as humorous sarcasm --Hammersoft (talk) 17:05, 11 April 2008 (UTC)

• • This is part of the problem. Editors wanting to use an image have to carefully explain how it meets ten different criteria. That is fine - I have no problem with that. Why, though, do those demanding an image be removed not have to do the same? They should have to list the ten criteria and say clearly which ones it fails. Instead, most non-free image admins I see say "blatant" this and "obvious" that, and then debate one or two points, instead of addressing each criterion one-by-one. I know that failing even one criterion is sufficient to have an image removed, but it would feel much more like people were being methodical and rational if they had to fill out a ten-point template when reviewing an image. Both those wanting to use it and those wanting to remove it. Carcharoth (talk) 17:07, 11 April 2008 (UTC)
    • Oddly enough, where there are borderline cases, I've noticed that I tend to end up having civil discussions with other editors that usually lead to a mutually acceptable solutions (or at least an ecceptance of why the policy is stated how it is). It's the removal of "blatant" failures of NFCC (i.e. huge decorative fair use galleries, like that article above) that causes more problems... Black Kite 17:12, 11 April 2008 (UTC)
      • What I find strange is that people used WP:IfD for a while, for discussions of this, and then stopped. IfD is rather a strange place. Seems to be orphan central. Carcharoth (talk) 17:19, 11 April 2008 (UTC)
        • Probably because people like me and User:Abu badali didn't want to deal with the stress of it any more. From August to about December 2007 I did almost nothing except POTD. howcheng {chat} 17:42, 11 April 2008 (UTC)

I support the retention of the original, more restrictive wording. As it is, we are the most liberal in terms of image use of all the various Wikipedia projects, as I understand it. Over time, I think it is inevitable that "fair use" images will be more stringently scrutinised, which is as it should be in a project which aspires to providing a resource which is free. Carcharoth, that sounds like a good suggestion and maybe there is some mileage in it, but it seems to contradict our principle of "when in doubt, leave it out". Certainly any proposal to change something fundamental like this should be able to demonstrate a far wider consensus than this one has. This may be of interest to those who have not seen it. --John (talk) 17:18, 11 April 2008 (UTC)

The vegan dinner. Yes. I've seen that. Carcharoth (talk) 17:20, 11 April 2008 (UTC)

Argh, I've been trying to find that since yesterday, I couldn't remember whose userspace it was in, and I was putting "vegetarian potluck" into the search. *slaps head* Black Kite 17:31, 11 April 2008 (UTC)

There are two contradictory uses of "'free"': all the content that other people can freely use, as compared to all the content that we can freely put into WP. What we should pick depends upon our primary goal: to be a source of material for other people to use, or to be an encyclopedia. I always thought it was an encyclopedia, and that this implied we wanted the best content we could legally use. So I obviously want the broadest interpretation of foundation policy that the words will bear, for that will give the most informative encyclopedia. DGG (talk) 18:23, 11 April 2008 (UTC)

• See Angr's essay. The idea that we are first and foremost an encyclopedia, and to heck with restrictions on fair use, is false. Just because we can use fair use doesn't mean we should. We aren't an encyclopedia. We are a free encyclopedia. See m:mission. See m:vision. You just can't be those things and still paint a broad paintbrush of what is acceptable fair use. --Hammersoft (talk) 18:34, 11 April 2008 (UTC)
  • Hammersoft is right; we are both an encyclopedia and and free content source. Both goals are equally important. howcheng {chat} 18:40, 11 April 2008 (UTC)

• ...and they aren't mutually exclusive in any respect. --Hammersoft (talk) 19:10, 11 April 2008 (UTC)

Angr's essay misses the point. If you're a vegan, the whole point is to have vegan content only. But en-WP is different. Our aim here is to encourage the creation of new, useful free content. That is why we do not accept non-free content that is rivalrous to the creation of comparable free content. But unlike the vegans, en-WP is not, and has never set out to be, a free-content-only encyclopedia. Where non-free-content is (i) legal, (ii) useful, and (iii) not in competition with free content, there is no reason not to include it. Jheald (talk) 10:13, 12 April 2008 (UTC)

• Wrong. Please see our policy at WP:NFCC. Non-free content has to jump through alot more hopes to be acceptable here than you describe. --Hammersoft (talk) 23:08, 12 April 2008 (UTC)

• The point of NFCC is to defend those three criteria above. I'd be interested to see you cite an example of an NFCC criterion which you think is not there to defend one of those three points, because that would be prima facie reason to relax it. Jheald (talk) 08:12, 13 April 2008 (UTC)

• This is certainly a valid and easier way to think about it than the 10-point approach, and, crucially, is still in the spirit of the Foundation Resolution. The main objection outside of the three points Jheald raises is the m:mission, which some people chose to interpret as "Free Wikipedia from fair use" and "Say NO to fair use". See my spectrum section below. Getting down to specifics, let's compare the three criterion approach to the 10 criteria in the NFCC:

• NFCC#1 - "replaceability" is covered by point (iii) "not in competition with free content". See also m:mission.
• NFCC#2 - "commercial opportunities" is covered by (i) "legal" (in a broader sense than "fair use legal"). Nothing at all to do with m:mission.
• NFCC#3 - "minimal use" is also covered by (i) "legal" (again, similar to but not exactly the same as "fair use legal"). See also m:mission.
• NFCC#4 - "previous publication" is also covered by (i) "legal" (though there are interesting points that could be discussed here). Nothing at all to do with m:mission.
• NFCC#5 - "content and being encyclopedic" (an internal rule) is covered by (ii) "useful" (though the word "useful" is a simplistic expression of this concept). See also m:mission.
• NFCC#6 - "Image use policy" (another internal rule). Not really anything to do with the m:mission, but aspects of the image use policy do cover other legal aspects of image use, so this partly speaks to (i) "legal".
• NFCC#7 - "no orphans" (really just an internal rule to avoid becoming an image hosting service, but has legal aspects as well - you have to use image if you want to claim fair use). In effect, this restates some legal aspects covered by (iii) "legal", but has little to do with m:mission.
• NFCC#8 - "significance". Partly related to m:mission as a tool to restrict excessive fair-use, but also related to an image being encyclopedic, so this falls under (ii) "useful". See also m:mission.
• NFCC#9 - "only in articles". Again, this is covered by (i) "legal", as fair use would only relate to articles (at least as far as Wikipedia is concerned - it quite rightly doesn't want to worry about defending uses outside of articles, and so just bans them). This could be seen as an aspect of m:mission, or rather reminding people what this is all about.
• NFCC#10 - "source, copyright tag, naming the article and writing a rationale" are all (to varying extents) legal aspects, and covered under (i) "legal". Not strictly related to m:mission, but more paperwork, though very important paperwork if you ever have to establish a legal defense.

So I agree that jheald's three-pronged approach would work, but most of the NFCC criteria unpack from the "legal" aspects, and because NFCC goes beyond US fair use laws in many aspects (NFCC is more restrictive), we need to unpackage all the various legal aspects and lay them out clearly. Having said that, it might be useful to clearly state which of the criteria are "legal", which are "mission" (free content) and which are "content" (encyclopedia). Carcharoth (talk) 09:14, 13 April 2008 (UTC)

Jheald, your opinion is factually wrong. You state that the point of the policy is to allow fair use where it is "(i) legal, (ii) useful, and (iii) not in competition with free content" This is incorrect. For beginners, it approaches from the wrong direction. The point of the policy isn't to allow fair use. The point of the policy is to restrict fair use. This is a fundamentally different approach and reading it from your perspective yields an entirely different understanding that completely misses the mission and vision of the Foundation. In fact, whether or not something is legal is really a very minor point. The policy is such a massive superset of actual law that if a fair use image truly, legitimately met our policy (most don't), they'd be well clear of any legal concerns. Whether or not something is legal is really a very minor consideration. Whether or not something is useful is only partially accurate. It's far, far from enough to be 'useful'. Further, whether something competes with free content isn't really all that strong of a point either, but it's probably the best point you've got going among these three points. Certainly a fair use image of a living person competes with a freely obtainable image of the same person. But this 'compete' stance lends strength to the (failed) idea that some people have that a clearly higher quality fair use image is acceptable in lieu of the absence of a good quality free content image. You've really missed the mark here.

In essence, you've entirely missed the point of the policy. Your interpretation of the policy would leave us believing that articles such as List of Kanon characters are perfectly acceptable locations for dozens of fair use images. Note this version of that article, which contains 22 such images, vs this version, which has two images. Your interpretation would say the 22 are ok; they're legal, useful, and don't compete with free content. That version of the article very clearly meets your criteria. Unfortunately, your criteria are quite wrong.

Of course, I don't expect you to be swayed by this argument. You'll disagree with it. You have always interpreted fair use policy very liberally and want as much fair use as you can get away with. Thankfully, the strength of the Foundation's mission, vision, and resolution stand behind those people who are trying to keep this as free a resource as possible, not a resource with as much fair use we can get away with. --Hammersoft (talk) 00:31, 14 April 2008 (UTC)

Fortunately we now have some quite detailed criteria for images in list articles. And note the tone -- they are positive and encouraging about how to use fair-use content appropriately, not taking as their aim how to hammer people down with restrictions.

As I understand it, those list-image criteria are drawn to ensure that anyone reusing the page in substantially the same form - even if in a commercial context - should be unquestionably within the U.S. fair use limits. I think they're probably about right for that purpose, but I am interested if you think they are unnecessarily restrictive vis-a-vis the legal position for those reusers. I wonder if you can produce evidence to support that, because (as I said above) if so I think that would be prima facie reason to review them.

Finally, I think you should be careful not to confuse whatever may be your own personal mission, vision and resolution with that of the Foundation. Because I see the Foundation statements as supporting a free-content mission, not a free-content-only mission; compatible with WP as the Foundation seems entirely relaxed for it to continue. Jheald (talk) 17:06, 14 April 2008 (UTC)

• I think you should be careful about conflating your own personal desires vis-a-vis non-free content and the Foundation's principals. I stand in support of the Foundation's edicts, and act wholly in support of them. Not ONCE have I ever crossed outside of that line. If you've got an axe to grind with me, out with it now. Otherwise, enough of the attacks upon me. --Hammersoft (talk) 17:20, 14 April 2008 (UTC)

• I am merely making the point that the personal interpretation you have adopted of the Foundation's edicts is not the only interpretation, nor I would submit the most plausible interpretation. Jheald (talk) 17:24, 14 April 2008 (UTC)

• And doing so by commenting on me rather than points raised. I find your behavior towards me today to be most reprehensible, and I strongly, strongly urge you to read and abide by Wikipedia:No personal attacks. If you can't do that, you might consider taking a break from making any comments towards me until you can. --Hammersoft (talk) 17:26, 14 April 2008 (UTC)

• And by reacting strongly to Jheald, you have also diverted discussion. I suggest both of you go back to the previous point in your discussion, and try to work towards useful criteria that can be applied in practice. Carcharoth (talk) 11:50, 15 April 2008 (UTC)

• --Hammersoft (talk) 13:12, 16 April 2008 (UTC)

Even though the Foundation has taken a strong stance that fair use should be done minimally, within narrow limits and that policy can not be eroded, the inclusionists insist on liberal inclusion.

Actually, the Foundation Policy says that we can use non-free content as long as it's part of our EDP. We should return to an EDP that allows content with permission of the copyright holder (or non-commercial use only, etc.), and then most of the fair use issues will go away. — Omegatron (talk) 23:09, 21 April 2008 (UTC)

Best practice examples for NFCC#8

How about another attempt at a (deliberately) long list of good example of NFCC#8. No better way to teach people than by showing them. I'll start the list off, and leave other to fill it out. Carcharoth (talk) 18:13, 11 April 2008 (UTC)

List of NFCC#8 examples

Article	Image(s)
More Demi Moore	1
Leibovitz v. Paramount Pictures Corp.	1, 2
Demi's Birthday Suit	1
Raising the Flag on Iwo Jima	1
1968 Olympics Black Power salute	1
Birmingham campaign	1, 2
Che Guevara (photo)	1
Steinberg v. Columbia Pictures Industries, Inc.	1, 2
Person of the Century	1

Please add more above and update the timestamp at right. Last updated (~~~~~): 16:00, 12 April 2008 (UTC)

I removed Billy Ripken because it's not clear it's an unambiguously good example of an image meeting NFCC#8. The sentence, "In 1989, Ripken's Fleer card showed the player holding a bat with the expletive fuck face written in plain view on the knob of the bat", if it were sourced (it isn't), would be sufficient to identify what we're talking about. The image itself is superfluous. —Angr 19:02, 11 April 2008 (UTC)

The image is hardly "superfluous." There is no way text con convey how visible the phrase is, or what its impact is in the context of a baseball card, or how the expression on Ripken's face relates to the phrase and its overall effect. An image of the card here provides invaluable information—information that is essentially visual. It is, indeed, an unambiguously good example of an image meeting NFCC#8. It's the misuse of NFCC#8 against such informative images that has made its rewording imperative.—DCGeist (talk) 19:15, 11 April 2008 (UTC)

How visible the phrase is is made clear by the efforts that were made to cover it up, which are described by the text. Any relationship between the expression on Ripken's face and the phrase would have to be explained in words (as it is, the article discusses no such relationship, nor is any apparent from looking at the card). The image of the card does not significantly increase readers' understanding of the article, nor would removing it be detrimental to understanding the article. It's the misuse of NFCC#8 in support of such useless images that makes a tightening of its wording to something along the lines of "A nonfree image may be used only if the article would be utterly incomprehensible without it" imperative. —Angr 19:53, 11 April 2008 (UTC)

No--how visible the phrase is is suggested by the text; it is made clear by the image. And no, the relationship between Ripken's expression and the phrase does not "have to be explained in words". It is precisely the sort of subtle visual information that requires an image to convey. And no, you can't simply ignore the overall impact of seeing the card and pretend that that is not significant information. It's an essential part of why we use media at all in this encyclopedia--because the sensual effect of many images and sounds cannot be conveyed sufficiently via text. It's the lack of understanding of how visual information works on the part of many editors who take it upon themselves to police our media usage that makes these matters so contentious.—DCGeist (talk) 22:24, 11 April 2008 (UTC)

I agree with Angr here. So how about compiling a list of borderline cases (the problem is using deleted images as an example - we would need to link to external websites legitimately using the photos)? Someone could trawl IfD for those - back when IfD was being used for such debates. It is clear, though the policies don't use such language and arguably should never use such language, that in practice there are "strong" and "weak" non-free uses. The classic case of "strong" non-free use is on articles about an image or about the impact an image had. All the "covers" that are not explicitly discussed in the articles about the objects, but are merely identifying (not decorative) images, could be described as "weak" (eg. album covers, book covers, magazine covers, cereal box covers, etc). Then you have the class of "poor" to "bad" fair use (these are ones that should be removed. Ones like decorative use of decorative galleries (though please note that galleries are acceptable if the layout of an article is better with a gallery rather than left or right-justified images), or individual illustrations in list articles. I'm not going to mention episode screenshots, as I don't know enough about them. Finally (though there are other uses not covered in this brief survey, such as historic pictures), there is the case of non-commercial pictures (some argue these can be uses under fair uses). On top of all that, pictures of uncertain copyright status (usually old B&W ones where no photographer was identified or the information has been lost) are often, in my opinion harshly, shown the door because of "doubt" - whereas I think we should exercise a judgement call and, providing the images meet all the other criteria, use them under non-free use until they become unambiguously public domain, and then switch the license, and if anyone claims copyright in the meantime, we update the image information accordingly and possibly take the image down if need be. A good example is the situation described here. The classic Chamberlain picture (which has an interesting history) is copyright Getty Images (I linked to a different source for convenience), but the whole debate was short-circuited by someone finding a suitable free replacement, namely: Image:MunichAgreement .jpg. One day, if someone can arrange it, a high-res scan of that image should be possible to obtain for Commons to update the current low-res picture. Here, the balance is whether having a free alternative outweighs the stuff that has been written about the more widely-known image (most of the stuff doesn't talk about the image, but more about the events of that day). In future, I can see the free image becoming more widely known if Wikipedia use it instead of the other one. But more probably the other one will become public domain at some point (though I've still to get a clear answer from anyone as to when the Getty Image will become public domain - if no-one bothers to work it out, or challenge Getty, then Getty might hang on to it for a good 10 or 15 years longer than they should...). Right, I went horrendously off-track there, but hopefully that one example might convince people that a showcase of IfD debates may help consolidate opinion and consensus on what the policies mean. If the last year was about NFCC#10c, maybe this year we can finally nail down what NFCC#8 means, and bring some stability to this area? Carcharoth (talk) 20:22, 11 April 2008 (UTC)

As Carcharoth is mildly putting it, the Guevara image is not public domain. I have been waiting for a little while for commons to make a decision about the copies there (!) before working on addressing the licensing tags on the copies that are here. — Carl (CBM · talk) 19:06, 11 April 2008 (UTC)

It would help to explicitly tell which image is being illustrated; most of the images at Birmingham_campaign are free, for example. — Carl (CBM · talk) 19:09, 11 April 2008 (UTC)

I linked to the images I think are referenced, except for "Manos" The Hands of Fate, where I can't tell which image is intended. — Carl (CBM · talk) 19:51, 11 April 2008 (UTC)

The one I meant there was the one showing the clapperboard actually in shot, showing the terrible production values. Having thought about that one, though, I suppose that could technically be described in text (NFCC#1). Black Kite 13:19, 12 April 2008 (UTC)

Oh, I hadn't realised the Che Guevara image was not PD. I must go and read the Commons debate. Linked a few more pics in the examples. I can't work out which of the two in "Manos" The Hands of Fate is meant either. Neither seem like great examples to me, in fact, I've removed it from the list for now. Carcharoth (talk) 20:34, 11 April 2008 (UTC)

Interestingly Image:Leibovitz-DemiMoore-VanityFair-150px.gif probably distorts the picture so much (very poor quality scan) that we are misrepresenting the original image! If we are going to use the image, we should use better quality ones like Image:Vanity Fair August 1991.JPG. Carcharoth (talk) 20:39, 11 April 2008 (UTC)

Examples of montage images

Strictly speaking, this is NFCC#3a, but since we are on a roll... Carcharoth (talk) 20:43, 11 April 2008 (UTC)

Article	Image(s)
List of characters in The Simpsons	1
List of The Batman villains	1
Great Books of the Western World	1
The Culture	1
Collaborations by Óttarsson	1

The first one is fine. The source is fox. The second one is not, and fails. Reason; it's a user created montage. This would be no different than creating a montage from all 22 images on List of Kanon characters and saying "Look ma! It's just one image!" It isn't. It's still 22. The second should be deleted. --Hammersoft (talk) 20:51, 11 April 2008 (UTC)

I think the argument presented on the second image is that it's a piece of artwork that was created in an 3:1 aspect ratio and then scrolled to fit onto 4:3. Sceptre ^(talk) 21:00, 11 April 2008 (UTC)

• That's not what the image description page says. It says the source is "Apparent composit from multiple screen captures". --Hammersoft (talk) 21:06, 11 April 2008 (UTC)

• I think that what happened is that the camera angle in the animation panned across the scene we see here, and the uploader took three screenshots in close succession and stitched them together. An equivalent would be to take a three second clip from the source (in this case The Batman episode "Rumors"). That would be truer to the source than the current user-generated composite, though a video clip, ironically, might be less palatable to the copyright holder (ie. legal exposure). It's an interesting example though. Another example of montages is people spreading out a collection of books or other products, with their covers visible, and photographing them for a list. For example, Image:BarnesNobleClassics.jpg at Barnes & Noble Classics Collection#Current Classics collection (should that be a GFDL license?) compared to Image:Call of the wild.jpg (wrong license) at Barnes & Noble Classics Collection#Children's hardcover collection. That's still not quite what I mean by a montage of book covers (by which I mean a single photo of a physically arranged collection, as opposed to a digitally produced montage of several separate book cover scans). Oh, here is a good example: Bibliotheca Teubneriana containing Image:Teubner covers Gk.jpg. Once cover is PD, the other two are copyright, but to be frank, deathly boring. Let try again (ugh, Category:Series of books contains some really rubbish articles). OK, I failed, but look at Gateway to the Great Books. Some would argue no image is needed at all, some would argue that the single image there is OK, and some would argue that you could spread the 10 volumes out on the floor and take a single picture to illustrate the article (or maybe a picture of the spines forming the 10-volume set, as seen at Great Books of the Western World for Image:Great Books.jpg). Another argument is that when the first edition of any book or magazine is old enough to be PD, that is sufficient to use on the article, and there have to be very good reasons to use anything else. Thoughts? Carcharoth (talk) 21:43, 11 April 2008 (UTC) Oh dear, now I want to buy Great Books of the Western World! And why are there so many redlinks there?

I agree about the Batman image. It appears to be a single image that was split up by its source, and then screencapped and reassembled by a viewer. -Freekee (talk) 16:56, 12 April 2008 (UTC)

How about Image:All The Culture Novels.jpg? --John (talk) 21:57, 11 April 2008 (UTC)

Perfect. Thanks. Goodness, I missed the fact that Banks has publised another Culture novel. Where's that wikibreak template... :-) Carcharoth (talk) 22:02, 11 April 2008 (UTC)

I still haven't read it myself, so I can't say anything about it. The image, while well-intentioned, certainly seems to raise some interesting questions though. --John (talk) 22:09, 11 April 2008 (UTC)

Another problem with montages is that quite often get tagged as self-made; thus. Black Kite 16:50, 12 April 2008 (UTC)

Yes, though that is a good example - I will add it to the table though it should be replaced with a list example. There has been artistic layout of the content of the picture, so the GFDL tag should remain, but there also needs to be a copyright tag as well. In many of these cases, the GFDL component of the photograph being taken is implicitly assumed or taken to be a "mechanical copy". This is clearly not a mechanical copy. Similarly, for Image:Great Books.jpg, the photographer has decided on an angle at which to take the photograph, and has considered the lighting and the arrangement of the books. Ditto for Image:All The Culture Novels.jpg - the choice to lay the books out in an arc shape is clearly one that the photographer made. What do people think about these aspects of things? Carcharoth (talk) 09:39, 13 April 2008 (UTC)

Aside: this may be the seeding of an NFCC#8 tutorial

I'm sorry to interrupt this excellent discourse with such a big visual intrusion, but I want to make the point that these borderline examples really do start to open up the issues for non-experts such as me. And it's non-experts who need to acquire more skill and knowledge on NFCC#8, because that will make your job of policing images much easier—less resistance when you do remove content, and more self-policing.

The trainer in me sees these examples and your accompanying debate as fodder for constructing a tutorial on the topic for non-experts. The wording of 8 needs to be vastly improved (involving some expansion, I fear), but the meat of the matter for non-experts will always lie in examples of what is only just acceptable and what just fails to be acceptable. You know the kind of thing I mean: present them with a succession of images; for each one, provide a little information—a few teasers—and ask them to judge whether it passes or fails NFCC#8. They hit the "Show" button, and the considered answer and rationale is displayed. Supply summary text at strategic intervals, driving home what they've learnt in the previous exercises. Possibly 10 or 15 images would do the trick. I've found this process to be very popular in providing opportunities for WPs to improve their editing skills. THIS PAGE is work in progress and has not yet been made available to the community; I really don't care for self-promotion, so I've linked it only to show you the possibilities for images (and indeed for sound clips, a major issue that passes you all by every day at WP:FAC—and which bothers me).

A tutorial, if people thought it was a good idea, could come only later, after a more effective policy and supporting wording has settled. It might be the opportunity to make 8 a truly effective and knowable pillar of NFC policy that more people in the community are able to embrace.

Please let me know if this is pie in the sky. It would, of course, require collaboration, since I have zilch expertise in NFC (which kind of underlines my point). TONY (talk) 03:01, 12 April 2008 (UTC)

This is exactly the sort of thing I would love to help out with and even write large part of if I got the time. I am most definitely not a lawyer, and I don't think anyone should need to be, but I agree that examples of past and present Wikipedia practice would help a lot. I would need help though. As WP:NFCC-C (my most ambitious attempt in this area, and one that has stalled for various reasons)) shows, it takes time, help from others, and the right sort of content, organisation and layout, for something to catch on and succeed. Much like the rest of Wikipedia, really. Carcharoth (talk) 09:58, 12 April 2008 (UTC)

And look HERE at how the folks over at the Featured picture criteria clarify by a link from their legislation what is and isn't acceptable technically WRT to their criteria. If they can do it, why can't we? TONY (talk) 10:20, 17 April 2008 (UTC)

One problem would be that we wouldn't be able to actually show the images on a policy page, which I believe would be necessary to depict the usage of the nonfree content in context. Perhaps if we can find some free imagery that serves the same purpose and tell the user "imagine that this is non-free... in this case, it would be unacceptable because ..." howcheng {chat} 16:45, 18 April 2008 (UTC)

Old IfD debates

Adding some links to examples of old IfD debates:

• 2007 July 18 (image - Max Planck Medal - deletion review)
• 2007 July 24 (image - Person of the Century)
• 2007 August 18 (image - Eugene O'Neill - more discussion - deletion review)
• 2007 August 20 (image - Munich Agreement)
• 2007 August 20 (image - Anschluss)
• 2007 August 20 (image - Toni Sailer)
• 2007 August 21 (image - Stanley Baldwin)
• 2007 August 21 (image - Neville Chamberlain
• 2007 August 23 (image - Victor Francis Hess)

Please add more if you think they are helpful. Please try to restrict to edge cases and long and useful debates. Maybe some shorter examples would be helpful as well? Carcharoth (talk) 15:58, 12 April 2008 (UTC)

• 2007 May 7 (image - Battle of Dien Bien Phu specifically this revision - deletion review - 2nd IFD)

Helps who to understand?

Images can help the understanding of those who can not read, of those who read English poorly, of those who are too young to understand all the words in the text, of those who have learning disabilities such they are far better at mentally processing visual imagery. Think of the children :) Help the disabled! Stand up for people who primarily communicate in other languages! Are we trying to be useful or elitist? WAS 4.250 (talk) 08:52, 13 April 2008 (UTC)

Yes, and going further there is the visual-verbal psychology thing. We must have an article on it by now! Theory of multiple intelligences is not really what I was looking for, unfortunately, and even spatial-temporal reasoning and visual thinking are in very poor shape. Something like this might be what I am trying to get at, but what WAS 4.250 said above is a more practical way of considering the impact of the use of visual media. Obviously it is not meant to be an excuse to put an image in every article, but it is making the point that some people find images more useful than others, so it is valid question whose views should be given more weight in this debate - those who are "visual types" or those or are "verbal types". Carcharoth (talk) 09:30, 13 April 2008 (UTC)

On the other hand (and as a trained teacher I really do understand the point WAS is making), we are a text-based encyclopedia, not a picture book and not designed for or aimed primarily at children. For the disabled or those who don't understand much English, there is always Simple. --John (talk) 16:58, 13 April 2008 (UTC)

The spectrum of opinions

The distinction needs to be drawn between people who clearly state:

• (1) Let's get rid of all (or nearly all) fair use (but only campaign for that at policy level).
• (2) Let's chip away at all cases of fair use until we approach as close as possible to no fair use (people campaigning on the ground to subvert the existing NFCC policy or promote the m:mission depending on your perspective).
• (3) Let's be very strict on fair use and only allow a few uses (working within policy).
• (4) Let's be very strict on fair use and only allow some uses (working within policy).
• (5) Let's be very strict on fair use on a case-by-case basis (working within policy).

Is that a fair characterisation of the various "strong" positions? The middle and "weak" and "liberal fair use" parts of the spectrum should also be described as well. Carcharoth (talk) 09:15, 13 April 2008 (UTC)

• I'd be surprised if there are that many people in (1) or (2). I consider myself very strict on fair-use but I still definitely fall into one of (3)-(5). Black Kite 09:39, 13 April 2008 (UTC)
  • I discussed this with Angr recently, and Angr openly admitted to chipping away with the larger goal in mind. My position actually varies on the type of content. I really can't bring myself to defend some cases of fair use, but I recognise that this is a case of "I don't like that sort of thing", and try to avoid those areas. In other cases, I think broad generalisations are possible, but mostly I favour working on a case-by-case basis, while recognising that there is a "volume problem". Carcharoth (talk) 09:45, 13 April 2008 (UTC)
    • Another problem is that most blatant violations of fair-use seem to, unsurprisingly, be in popular culture articles (see CAT:NFIO), which can leave editors working in this area open to charges of WP:YOUJUSTDONTLIKEIT. Black Kite 10:00, 13 April 2008 (UTC)

Kite, since you raise the matter of pop culture articles, as an FAC reviewer, I'd love to know how strict we should be on the 30 s sound-bites. Reviewers hardly ever say anything; I sometimes question nominators about the claimed "educational" value where nothing much is said about the specific audio clip in the adjacent main text. Comments? TONY (talk) 13:33, 13 April 2008 (UTC)

That's a tricky one. Technically, under minimal use, we shouldn't be using any, but I totally agree with the legitimacy of sound samples where they show something which is heavily discussed and important to the article, such as a new style of music, or something ground-breaking or unusual which it would be too difficult to describe in text. Or on an article on a music genre, a representative sample or two. What we shouldn't have is articles with a clip from every single or album, where it doesn't assist the reader. They're just effectively fair-use image galleries. Recent examples - look at this [1] or [[2]]. Black Kite 13:42, 13 April 2008 (UTC)

In which case, if you don't mind, I'll alert you next time I'm suspicious at WP:FAC ... TONY(talk) 14:22, 13 April 2008 (UTC)

No problem. Black Kite 15:33, 13 April 2008 (UTC)

I personally fall into (3)-(4). As far as sound clips go, I go down to about (2): I only upload sound clips iff I can write at least one decent-sized (60-80 words) paragraph about it. Sceptre ^(talk) 14:47, 13 April 2008 (UTC)

I am somewhere in 3-5, My position is remove the image if you cannot make a solid defense for keeping the non-free media that is more than just "I LIKE IT". As the usage and defense of non-free media falls on those who want to include the media. If you cannot defend the usage it shouldbe removed. This may shock some people, but there are some non-free content usage that I do defend. β^{_{command 2}} 15:48, 13 April 2008 (UTC)

• (6) We should avoid fair use unless absolutely necessary, but allow other non-free, legal, informative content instead, with permission of the copyright holder. Fair use would be permitted only if the copyright holder refuses permission or cannot be contacted, and no equivalent works under a free or more permissive license can be found.
  • Minority position so far.  :) — Omegatron (talk) 23:36, 21 April 2008 (UTC)

NFCC8 and disruptive use

Can we put the point across that if you remove an image per NFCC8, you better be prepared to explain why? I think the whole "NFCC work is automatically undisruptive" idea that's floating through people's heads is wrong: over at the Doctor Who project, we've had users like Fasach Nua (talk · contribs) and Bragen (talk · contribs) on single-purpose fair use crusades just on those articles, and no-one can say it wouldn't be disruptive were they free. Sceptre ^(talk) 00:37, 14 April 2008 (UTC)

• We already had the screenshot-per-episode debate a long, long time ago. Fasach Nua and Bragen happen to be right. --Hammersoft (talk) 00:58, 14 April 2008 (UTC)

  Whether they're right is superficial. This is about disruption. Wikipedia:Requests for arbitration/Episodes and characters 2 pretty much made the two seperate. Sceptre ^(talk) 01:09, 14 April 2008 (UTC)

• This is reverse thinking. The people policing fair use overuse aren't the ones that need to be jumping through major hoops to explain why the image is being removed. The people who want the fair use are the ones that should be doing the explaining. For example, Bragen removed the image Image:Smith and Jones.jpg from Smith and Jones (Doctor Who) [3]. The caption of the image referred to a genetic transfer. At the time the image was removed, the plot summary didn't even reference this. I.e., the picture was effectively useless to the plot description. Even now, the image description page just says the purpose of hte image is "To illustrate a key element of the plot". This is a seriously lacking fair use rationale. Not to mention the fact that two people kissing each other can readily be imagined by people; the removal of the fair use image would not in any way be detrimental to the reader's understanding of the episode. The fair use image brings nothing to the table that text saying "The Doctor performs a "genetic transfer" - kissing Martha to obtain human DNA" which the article now says. The fair use image is useless in this context, and I'm removing it. --Hammersoft (talk) 13:57, 14 April 2008 (UTC)

• I'm not arguing whether they're right or wrong. I agree that most of the Series 3 screenshots suck. This is about people using NFCC8 as carte blanche to go on single-purpose crusades. Sceptre ^(talk) 15:42, 14 April 2008 (UTC)

• It's not a crusade. It's supporting our policies and mission here. The correct response, rather than calling this a crusade, is for people who want these images included to come up with a detailed explanation why these images should remain. Looking around today, I'm finding all sorts of images that blatantly fail. They've been added just to illustrate the article. Take Image:Buffy 102 image.jpg for example. I can't even make out what Buffy's doing here, much less find any mention of it in the article. Removing an image like this isn't a crusade, it's just normal operations. People need to do considerably better work in including such images. We don't accept copyright violations here, and have carte blanche to remove them. Similarly, images which so obviously fail our non-free content guidelines suffer similar fates. --Hammersoft (talk) 15:51, 14 April 2008 (UTC)

• For the last time, I am not debating the worth of the images themselves. Have a serious look at the contributions of both users, and tell me that isn't single-purpose editing. Neither user has removed any fair-use image outside the Doctor Who scope. They don't bother discussing either, instead they engage in drive-by tagging. We block users routinely for tendentiously editing on a small area of articles, so why not fair use? Being right does not equal being non-disruptive. Sceptre ^(talk) 16:02, 14 April 2008 (UTC)

• Editing within a particular area doesn't automatically make it disruptive. Further, Bragen has done removals of fair use images outside of Doctor Who. Please see example. So has Fasach Nuaexample. Your claim is inaccurate. I strongly recommend against blocking these editors, and instead discuss the issue with them. You've not edited Bragen's talk page at all, and haven't touched Fasach Nua's talk page since January. --Hammersoft (talk) 16:09, 14 April 2008 (UTC)

• This is why—I'm afraid to say it—more rather than less text is required, and a solid attempt should be made to open up the fine distinctions to the troops, to minimise the horrible fights that go on. The imprecision of Criterion 8 is a recipe for big trouble, and we need to make the job of our NFC people easier, not more difficult. TONY (talk) 02:21, 14 April 2008 (UTC)
• I think you'll find that those dealing with fair-use on a regular basis are quite happy to explain why images are removed - we have plenty of practice, given that it is often necessary to explain the same point multiple times on the same talk page, even when explaining to editors such extreme examples as why galleries of 20+ non-free images need to go! But again, this links back to #8's second clause - it is sometimes difficult for editors to understand why images need to go, because they think it makes the article "better", and they conflate that with "easier to understand". Black Kite 15:58, 14 April 2008 (UTC)

  I know, and that's good. People like you and Future Perfect have been helpful with Doctor Who images. I'm talking about the people who don't. Sceptre ^(talk) 16:02, 14 April 2008 (UTC)

• Further, this is why I think the wider problem is not the wording of #8 per se, but the fact that (a) there are so many articles that violate multiple parts of NFCC, and (b) (which leads on from this) that many editors don't even realise NFCC exists. The recent shenanigans with BCBot and so on haven't helped, because it's confused many editors into thinking that all non-free use is OK as long as it's got a fair-use rationale! These two things IMHO need to happen
  • We need to make NFCC, however it's worded, far more visible
  • The articles that seriously violate NFCC (CAT:NFIO, User:BKNFCC/List) need to be made compliant. Incidentally, if anyone's got a script to generate an up-to-date version of User:BKNFCC/List, it'd be marvellous, because that's hugely out of date, and when I go through checking it, 75% of those articles have been fixed (conversely I keep finding lots of violating articles that aren't on it too).

Black Kite 15:58, 14 April 2008 (UTC)

• • • I have to disagree: leaving open to subjective judgement the boundary between "significant" and "not significant", between what is detrimental and what is not, is asking for trouble. It's not only a matter of making the policing decisions easier to arrive at, but of guiding editors in their search for images and audio clips (waste of everyone's time in the first place not to provide explicit guidelines). TONY (talk) 09:59, 15 April 2008 (UTC)
      • Exactly my point, really; the second clause of #8 makes enforcement of flagrant violations much easier, but I agree it doesn't really help on single images that possibly fail #8 only. The trouble is, how do you formalise a word like "significant"? Even if you leave it out, there is still room for wikilawyering on both sides. Black Kite 12:49, 15 April 2008 (UTC)

The only way to do this is to provide examples of cases just either side of the boundary between significant and not significant. Examples are just what I managed to weed out of the criteria in my overhaul last May, because they were cluttering the text, making it harder for non-experts to comprehend, and were in some cases not well thought through. A footnote at the end of 8 might do the trick, leading to an explanation that just what is sufficient to satisfy the test of significance can most clearly be explained in a set of examples. There could be a link from that footnote to a dedicated page of examples, perhaps Wikipedia:Non-free content criteria/NFCC#8:The test of significance. One could start with a few examples that are obvious fails, such as your example of the ridiculous gallery of 20 no-nos; these could serve as warnings of what not to do, and because attached to an official policy criterion, would carry the kind of weight with prickly editors that would enable quick resolution of resistance and brush-fires via one easy-to-insert link at the talk page. Moving on, you'd have the more difficult borderline cases that would serve to define the boundary for both those who monitor and those who seek to upload in the first place. The test of significance would make everyone's job a lot easier and would help in recruiting more WPians to the task. What do you think? TONY (talk) 15:26, 15 April 2008 (UTC)

• In other words, an improved and illustrated version of User:BKNFCC/FAQ#FAQ_and_frequent_defenses? Yes, that'd be good. Black Kite 15:44, 15 April 2008 (UTC)
  • Not so much improved—since that page serves a different purpose and seems pretty good—but different in goal, structure and content. I'm wondering whether the examples people have provided of difficult/borderline cases are a starting point. Can some of those images be framed as only-just-a-fail and only-just-a-pass examples when put to the test of significance? Can some illustrate more obvious fails and passes of the test? If so, perhaps you could identify them and we could construct a few examples to get feedback here. It may be that the current wording of 8 can survive in its indefinite, fuzzy frame if supported by such a page. TONY (talk) 15:57, 15 April 2008 (UTC)

These instructive borderline examples need to be analyzed as well in the context of NFCC#1 and NFCC#3a.

In many cases, the significance of a fair use image is judged relative to a free one—argument erupts over whether the latter is sufficient to fulfill the encyclopedic purpose of the former. In such cases, NFCC#8 and NFCC#1 are and must be considered together.

In many cases, editors will argue on behalf of the significance of multiple fair use images in an article—each increasing readers' understanding by providing distinct information. Others will argue, in effect, that the minimal usage provision of NFCC#3a more or less trumps NFCC#8 and the number of images must be reduced, regardless of their individual significance. Again, if we're giving examples of borderline cases that are meant to instruct our editors, NFCC#8 cannot be considered in isolation from NFCC#3a.

As for the current wording of NFCC#8, it is still a detriment overall to our work here, even as it serves an important purpose, however crudely. In its current form (the result of the random timing of administrator Kwsn), it is a sledgehammer, when the debate above demonstrates that what we desperately need are a good set of scalpels. Defining what we mean by significance—in the light of both "free equivalence" and "minimal usage"—will give us the tools we need to do less harm and more good.—DCGeist (talk) 18:04, 15 April 2008 (UTC)

To stretch an analogy waaaay too far, if you're facing an army (in this case of thousands of non-compliant articles), a scalpel isn't going to be as useful as a sledgehammer. Black Kite 14:43, 16 April 2008 (UTC)

Agreed, and my increased appreciation of the problem out there and the effort involved in containing it has affected my position here. Ultimately, however, our image policy exists in order to help us balance our objectives of creating an encyclopedia as free as possible and as excellent as possible, which is the very heart of our mission.

And, by the way, I'm not opposed to the idea of you having a sledgehammer. It's just that something so deeply subjective as the currently worded criterion 8 is a very bad choice for it, because it allows that same sledgehammer to be misused by people who are a lot less trustworthy than you.—DCGeist (talk) 18:24, 16 April 2008 (UTC)

A sledgehammer operating in a smarter, more fine-grained regulatory environment would be better. It wouldn't have to be used as often, then. TONY (talk) 08:28, 18 April 2008 (UTC)

Entanglement of 8 with 3a

Dan Geist's point about the need to consider 8 in association with 1 and 3a led me to examine the wording of 3a.

Minimal usage. As few non-free content uses as possible are included in each article and in Wikipedia as a whole. Multiple items are not used if one will suffice; one is used only if necessary.

The last clause—"one is used only if necessary"—is totally redundant in the light of 8, and I think it should be removed. There's a case for retaining the previous clause—"Multiple items are not used if one will suffice"—since having established that one image passes the test of significance (8), a wilful editor might argue that a suite of similar images are just fine. But one versus none is fully covered in 8. TONY (talk) 14:17, 16 April 2008 (UTC)

PS In any case, the language sucks: "As few items of non-free content as possible are ...". TONY (talk) 14:18, 16 April 2008 (UTC)

• Strongly oppose that - it'll cause complete confusion. The last clause of 3a is there to emphasise that "minimal" can mean "none". If you remove it, it suggests that one is OK in all circumstances, regardless of how #8 is worded. Black Kite 14:40, 16 April 2008 (UTC)
• I see what you mean. But it has an awkward relationship with 8, doesn't it. Any ideas as to how this might be resolved? (And you have no objection to fixing the language in the first clause?) TONY (talk) 15:01, 16 April 2008 (UTC)

• I can't support removing that entriely. The role of that clause is to explain that "minimal use" might be zero nonfree images. Otherwise, the clause ends with "Multiple items are not used if one will suffice." That might be interpreted to mean that one image qualifies as minimal use. Maybe the last part can be rephrased, but I can't think of a good way right now. — Carl (CBM · talk) 15:05, 16 April 2008 (UTC)

What about this:

Minimal usage. As few items of non-free content as possible—if any—are included in each article, and as few as possible in Wikipedia as a whole.

I can't make out why the multiple–single–none point needs to be made now; it's stronger because the "none" point is now embedded more dramatically and earlier. And it resolves another problem I just noticed in the current wording: there's ambiguity as to whether the second sentence—"Multiple items are not used if one will suffice; one is used only if necessary."—refers to individual articles and the whole project, as explicit in the first sentence. Well, it can't can it, yet it's fuzzy at the moment. TONY (talk) 15:10, 16 April 2008 (UTC)

• Yes, that's better, but now I'm wondering if it's possible that somehow 3 and 8 can be conflated into one policy, to reduce the possibility of confusion even more. Black Kite 15:17, 16 April 2008 (UTC)

• OK, I'm glad you like it more. We could leave this lie for a little while as the messy debate on 8 matures, or you could inject this into that debate. I'm unsure. TONY (talk) 15:24, 16 April 2008 (UTC)

• • An important point here is the use of the words "necessary" (#3) and "significant" (#8). People have in the past argued endlessly over whether there is a need for this to be different, and what it all means, and whether one word is more restrictive than the other. Carcharoth (talk) 15:49, 16 April 2008 (UTC)
    • So getting rid of "necessary" in this proposal (see above) does resolves this issue? TONY (talk) 15:59, 16 April 2008 (UTC)
      • I think the idea is to first work out if it is necessary (that relates to NFCC#1 - if a free equivalent, and that includes textual description, is available, then it is not necessary to have even one non-free image), and then to work out if it is significant. Carcharoth (talk) 16:09, 16 April 2008 (UTC)

Yes. It would be clearer if the "necessary" test was restricted to crit 1: Is the image unnecessary because it (a) is replaceable by a free equivalent or (b) conveys no information beyond what can be conveyed by text?

That improvement would still leave a semantic problem with crit 3a: There remains an irresolvable tension between the terms "possible" (3a) and "significant" (8). Editor A argues fair use images X, Y, and Z each distinctly and significantly increase readers' understanding of article topic 1. Editor A makes a specific case for the significance of each. Editor B prefers not to debate the merits of the significance cases. Editor B argues that it is "possible" to use fewer fair use images; indeed, that it is "possible" to use none at all.

In sum, I believe crit 3a needs to be either fully recast or else integrated with crit 8. In terms of minimal usage within articles, what a stand-alone crit 3a can and should be helpful for is pointing us to the two specific ways FU image use can be reduced by informational tests that are not in conflict with significance: (i) reducing redundancy—two images should not be used when they both adequately convey the same significant information, and (ii) increasing efficiency—if there is one image that adequately conveys the significant information embodied in two different images, the single image is preferred. That could be suggested by the following:

Minimal usage. Multiple items are not used if one will suffice to convey the same information.

Beyond that, we can rephrase the first clause to emphasize the preference for a small quantity, possibly zero, while avoiding the highly problematic "possible," resulting in this:

Minimal usage. Few, if any, items of non-free content are to be used in each article. Multiple items are not used if one will suffice to convey the same information.

Upon consideration, the clause concerning Wikipedia as a whole does not belong in the criterion. It is superfluous and redundant—these are criteria for image use within articles; all the criteria to some degree serve the rationale of limiting the total amount of non-free content hosted by Wikipedia.—DCGeist (talk) 17:57, 16 April 2008 (UTC)

I don't think 'few, if any,' has the same meaning as 'as few as possible', and to the extent that they differ I think the latter is more in line with out mission.

The "Wikipedia as a whole" part does have some use; it applies especially to artwork such as Metamorphosis III and Regular Division of the Plane. Rather than repeating these images on every article that mentions them, we just put it on the article about the artwork, and link to that article from other places that mention it. The same is true for many works of art that have an article about them. Compare Guernica (painting), which is currently on several more articles than necessary (e.g. Suffering).

— Carl (CBM · talk) 19:33, 16 April 2008 (UTC)

Indeed, "few, if any" does not have the same meaning as "as few as possible." The latter inspires irresolvable conflicts with other criteria that the former does not. The meaning embodied in the phrase "as possible" that is in line with our mission may be more effectively expressed as part of the "replaceability" test of crit 1 and the "significance" test of crit 8.

Your second point, about not using images in more articles than those in which they serve a distinct educational purpose, (a) is not adequately conveyed by the current language of 3a and (b) is covered by crit 1 and crit 8. We agree, for instance, that Guernica does not belong in the article on suffering, but it is neither necessary nor helpful to call on crit 3a to determine this. As an aesthetic rendering of the experience of suffering, Guernica is readily replaceable by one of the many important artworks in the public domain that also clearly express suffering (crit 1). There is no discussion in the article indicating the particular significance of Guernica to the topic (crit 8). Rewording crit 3a along the lines suggested will make it more useful because its specific meaning and purpose will be much more clear.—DCGeist (talk) 20:05, 16 April 2008 (UTC)

After a night's sleep, I thought of this possible wording:

Minimal usage. Multiple nonfree items are not used if one will suffice to convey the same information. If the same information can be conveyed without the use of nonfree items, none should be used.

This restores that clause I am concerned about (to remind that zero is a real possibility) and removes the word necessary that was the source of the objection. — Carl (CBM · talk) 13:36, 17 April 2008 (UTC)

But Carl, the zero possibility is very plain from the "—if any—". I think your version is a substantial weakening of the guideline: I can see wikilawyer arguments over "the same information" (of course the images convey different info—they're different images). TONY (talk) 14:05, 17 April 2008 (UTC)

Weakening in which direction? I think it is between DCGeist's version and the current version in terms of how much nonfree usage it appears to permit. — Carl (CBM · talk) 14:41, 17 April 2008 (UTC)

I think Tony's perception is right here. The second clause you've suggested essentially repeats crit 1: don't use nonfree content where the same educational purpose can be fully satisified by free content. On the other hand, a concern about emphasis is legitimately placed here (in your words, "to remind that zero is a real possibility"). While, as Tony says, "if any" makes it plain, we can make it plainer. How about something like this:

Minimal usage. No more than a few items of non-free content are to be used in each article; none at all may meet the criteria. Multiple items are not used if one will suffice to convey the same information.

The order of the two clauses is easily reversed, making the point more conclusive, and perhaps slightly more emphatic:

Minimal usage. Multiple items of non-free content are not used if one will suffice to convey the same information. No more than a few non-free items are to be used in each article; none at all may meet the criteria.

The phrase "none at all may meet the criteria" is specifically inspired by the look I've taken at the work being done by administrators such as BlackKite involved in image policy enforcement, and is meant to be both clear and useful.—DCGeist (talk) 20:19, 17 April 2008 (UTC)

I'm not comfortable with the phrasing "No more than a few non-free items are to be used in each article," as it implies an absolute, though indeterminate, numerical limit for the number of items that can be used in an article, no matter how long the article or how many sub-topics it covers. While I have never been happy with the "as few as possible" wording, at least it leaves open the interpretation "as few as possible to convey the encyclopedic information that the items are intended to convey". The proposed rewording would seem to say that even if, say, twenty non-free items are absolutely vital for an understanding of a certain topic, and the items meet all the other NFC criteria, they are still not allowed by #3a, because twenty is more than "a few." DHowell (talk) 23:10, 17 April 2008 (UTC)

As has been true of everyone in this thread, DHowell makes some very good points. And, indeed, the word "few", while useful for indicating the general direction required by overall policy, is an open invitation to endless, irresolvable debate (yes, there will always be endless, irresolvable debates here...but we can stop inviting them so often). Is it possible to address all the main points that have been made while retaining the utility of the criterion and making it clearer and thus less prone to misuse? Perhaps something like this:

Minimal usage. Multiple items of non-free content are not used if one will suffice to convey the same information. There is no minimum number of non-free items to which all articles are entitled; for certain articles, no non-free content at all meet the criteria.

From the comments by Carl and BlackKite and from what I've observed going on in the trenches, it is unquestionably important to indicate that zero is a possibility—reinforced by the statement that there is no preset entitlement ("Well, that article has x# of fair use images, so ours gets x#!" No.)—DCGeist (talk) 07:47, 18 April 2008 (UTC)

1. Do we need to make an explicit point about the fact that there's no quota? Do some editors think this? How could they in the light of the criteria and policy as a whole?
2. In terms of the language, could I suggest a few things, because I think readers have to work too hard to understand it. Maybe end the first sentence at "suffice"? Uneasy about "minimum ... all ... certain ... no ... all". Does it mean "There is no automatic entitlement to use non-free content in an article"? Perhaps this might cover the second and third sentences. The shorter the better; I still like "—if any—", which is strong and punchy. TONY (talk) 08:25, 18 April 2008 (UTC)

I like "if any" too, but....

I think the phrase "to convey the same information" is important, to prevent the clause from being misused to demand "just one in the article" where that's inappropriate. It also helps, again, to focus on the practical objectives of reducing redundancy and increasing efficiency. Your second suggestion seems just right. Let's see where that would bring us:

Minimal usage. Multiple items of non-free content are not used if one will suffice to convey the same information. There is no automatic entitlement to use non-free content in an article.

Well. I believe that is both very clear and useful.—DCGeist (talk) 08:58, 18 April 2008 (UTC)

I'm still concerned that it's too easy for wilful editors to game the system by saying that "of course the second image doesn't convey the same information: it's not the same image", which will be true. TONY (talk) 09:13, 18 April 2008 (UTC)

Well then maybe the answer is to repeat the crucial term from criterion 8: "Multiple items of non-free content are not used if one will suffice to convey the same significant information."

On one side, we want to cut on the sort of gaming you describe; on the other, we want to cut down on false arguments for absolute numerical limits as DHowell describes (I think simply "Multiple items of non-free content are not used if one will suffice" would still cause lots of problems in practice—"Suffice for what?"). Productively, we want to encourage efficient (and thus less) use of non-free content—such as a group shots of characters in a TV series rather than individual shots of each. This wording should accomplish that; by using the exact same term as criterion 8, we should be able to keep it clear that the basic "significance" test is located there while clarifying the "efficiency" test that this criterion is practical for. It shows in a productive way how the two criteria relate while getting reducing the fuzzy entanglement that currently exists.

I realize now the wording could be a bit trimmer:

Minimal usage. Multiple items of non-free content are not used if one can convey the same significant information. There is no automatic entitlement to use non-free content in an article.

—DCGeist (talk) 18:37, 18 April 2008 (UTC)

That's acceptable to me. — Carl (CBM · talk) 18:51, 18 April 2008 (UTC)

What are thoughts on replacing "same" with "similar", "reasonably equivalent" or something else along those lines? To continue the cast example, a group shot of four characters is not the same as individual images of, say, five different characters; the group shot, however, would still be expected to convey satisfactorily the necessary information. ЭLСОВВОLД _talk 18:55, 18 April 2008 (UTC)

Well, I don't think so. If the argument that it is significant to illustrate each of the five characters succeeds on its merits, then a group shot including four does not satisfactorily replace individual images of all five. Per the clear language proposed, such a group shot must be used to replace individual images of the four characters represented therein, but the matter of the fifth would not and should not be covered by this particular criterion. Terms like "similar" or "reasonably equivalent" make the criterion less clear and thus less generally practical and effective.—DCGeist (talk) 04:06, 19 April 2008 (UTC)

• Dan, this is getting close. I do share elcobbola's concern about "the same", which pushes the argument too far in favour of wilful editors who'll argue that their multiple pics aren't the same. Similar at least allows those who monitor NFC more lattitude, even though it's less precise. In fact, it's the very precision of "the same" that will cause problems here, don't you think? I'm happy with "similar significant" or "reasonably equivalent, significant", although the second not such a neat expression. TONY (talk) 05:43, 19 April 2008 (UTC)

I was very excited to discover that we'd arrived at language that both Carl and I found agreeable. I see your point, Tony (and elcobbola): "the same" may well be too tight a term in practice. "Similar" strikes me, in turn, as just a bit too loose. Perhaps an adjective somewhere in between that both gives our enforcement administrators latitude and our (good) article developers clear direction: either "equivalent" or "comparable" might serve:

Minimal usage. Multiple items of non-free content are not used if one can convey equivalent significant information. There is no automatic entitlement to use non-free content in an article.

or

Minimal usage. Multiple items of non-free content are not used if one can convey comparable significant information. There is no automatic entitlement to use non-free content in an article.

I feel we are close, indeed.—DCGeist (talk) 07:05, 19 April 2008 (UTC)

Good, Dan. One item, so that "one" doesn't sound like the third-person pronoun. "Comparable" or "equivalent"—either; what do others think? TONY (talk) 10:36, 19 April 2008 (UTC)

Well, we seem close to being able to make a formal proposal, say for the following:

Minimal usage. Multiple items of non-free content are not used if one item can convey equivalent significant information. There is no automatic entitlement to use non-free content in an article.

The criterion would thus embody two elements:

1. A clear and practical test for both article developers and enforcement administrators: Can the significant information embodied in two or more non-free items be conveyed by just one item? If so, that single item must be used to the exclusion of the others.
2. A clear advisory for article developers and a practical tool for enforcement administrators: You are not entitled to use a certain amount of non-free content simply because other articles use that amount or more of non-free content; indeed, you are not automatically entitled to use any non-free content. All non-free content must meet the full set of criteria. For some articles, very little non-free content meets the criteria; for some articles, none at all.

Question: Should the proposal be made here, or the thread copied to the main NFC Talk page and the proposal made there? The proposal should obviously be advertised at whichever of the two pages does not host the ultimate discussion. Where else? WP:VPP, certainly. Wikipedia:AN?—DCGeist (talk) 21:22, 20 April 2008 (UTC)

• Probably post existing and proposed at NFC talk with a short rationale and link to this section, a declared intention to implement by (certainly before the end of the month), and an invitation to comment. TONY (talk) 02:19, 21 April 2008 (UTC)

• The proposal has formally been made at Wikipedia talk:Non-free content.—DCGeist (talk) 19:41, 22 April 2008 (UTC)

Avoid instruction creep

I'm seeing a lot of arguments about images which fail multiple criteria of the NFCC, and how #8 is entangled with #3a or #1 or other criteria. But if #8 only helps with images which also fail other criteria, then it is unnecessary instruction creep and should be eliminated. It has been made abundantly clear that a non-free image must pass all the criteria; an image need only fail one of the criteria to be deleted. For each of the other criteria, it is probably not to difficult to come up with an example which a consensus would agree fails that criterion alone: A current image of a living person could fail #1 yet pass #2-10; an image copyrighted by Encyclopedia Britannica would fail #2 and yet could pass #1 & #3-10; a single image could pass all 10 criteria but a second image which conveys the same information as the first would fail #3 (even though the same image could pass #3 if the first image were deleted); et cetera. If #8 is to have any meaning whatsoever in judging the use of non-free content, then we need to be able to clearly identify content or types of content that would fail #8 and #8 alone. The IfD debates which were posted above by Carcharoth don't help, because the images which were deleted were deleted for failing a criterion other than #8, and there was no consensus to delete the other images. NFCC #8, however it is worded, should have a clear purpose that is separate from, not entangled with, the other criteria. (This should apply to all the criteria, but since this discussion is supposed to be about #8, I'd like to keep it on that topic.) DHowell (talk) 05:12, 17 April 2008 (UTC)

I agree with this 100%. In different ways, this is the essential point I've been getting at in my recent observations: all the criteria should be worded so that the meaning and purpose of each is (more) clear and distinct, not (so) fuzzy and—the crucial term introduced by Tony in the preceding thread—entangled.—DCGeist (talk) 06:01, 17 April 2008 (UTC)

I must say that DHowell's statement makes a lot of sense. TONY (talk) 08:59, 17 April 2008 (UTC)

Oppose this, because there are cases where something can meet all the other criteria but still fail #8, especially when we have the "omission is detrimental" clause. Image:Cogny Castries Navarre.jpg is a good example of this, I believe. I'll put the relevant links in the above section. howcheng {chat} 16:31, 18 April 2008 (UTC)

I don't disbelieve you, but it's hard when the image can't be viewed. TONY (talk) 05:46, 19 April 2008 (UTC)

See [4]. howcheng {chat} 03:51, 20 April 2008 (UTC)

That is an interesting edge case in relation to its use in Battle of Dien Bien Phu, but it's not the best example because a bad decision was made to delete the image entirely. Obviously, the image should have been retained even if only for its use in René Cogny, where it was the sole image illustrating the now-deceased person who is the topic of the article. That particular photograph shows him engaged in the action for which he is most notable, as a commander of French forces during the Battle of Dien Bien Phu. It easily passes any reasonable interpretation of criterion 8 in the context of René Cogny.—DCGeist (talk) 03:47, 21 April 2008 (UTC)

I don't think this image is a good example for NFCC#8. The first IfD was overturned at deletion review, so I don't see much consensus that this failed #8. Even after NFCC#2 was brought up additionally in the second IfD, the consensus wasn't really all that clear, but I think the community is more tolerant of deletions based on NFCC#2 (where the risk of legal trouble is clearer) than deletions based on an alleged NFCC#8 violation. DHowell (talk) 05:34, 22 April 2008 (UTC)

A current case which might be relevant

See Stargate (device), its history and its talk page. Lots of images that skirt the fringes of 3a and 8 (though some clearly fail as well). Black Kite 11:31, 19 April 2008 (UTC)

OK, I'm interested to see how you deal with this one. Can you let us know when it happens? WRT the difficulty of using inadmissable content in a tutorial or explanatory link, the solution is to use free material but on the premise that it's non-free (for the purpose—to pretend, as it were). TONY (talk) 12:51, 19 April 2008 (UTC)

I applied the tag, and was hoping the discussion would come up with a consensus from regular editors on the article as to how the non-free content could be pared down. Unfortunately, that resulted in 3 of the 23 images being removed and the over-use tag being removed too. It wasn't helped by an admin on the talkpage who also doesn't appear to fully understand NFCC (which probably proves my point about accessibility and visibility). Anyway, I've posted links to the relevant policies, and we'll see what happens now. Black Kite 13:15, 19 April 2008 (UTC)

So it's not gonna be easy when there's a subjective element to how many are acceptable (3a and 8). Is the test of significance the logical way to satisfy 3a, in that order? TONY (talk) 13:40, 19 April 2008 (UTC)

I think in this case the main issue is #8 in that whether the images are necessary. There is also an element of #1 in this (are the images replaceable by text? Some certainly are) but of course there is the over-arching element of #3a - I would doubt if there is ever a rationale for having 20 non-free images in an article. Black Kite 15:19, 19 April 2008 (UTC)

I just nuked four, A few more need taken out. I know the subject fairly well given the detail and complexity of the subject it does need above average NFC images. β^{_{command 2}} 15:42, 19 April 2008 (UTC)

No, I realise it can't be demonstrated clearly with a couple of images, but 20 is clearly too many. Sgeureka has now posted a rationale for which ones should stay and go on the talkpage. Black Kite 11:48, 20 April 2008 (UTC)

It is clear from the discussion above that there is no consensus for the idea that the NFCC were intended to impose strict numerical limits to the number of images in an article. So would you please explain, to those of us who don't seem to see things quite so "clearly", using citations to clear statements from the Foundation, to discussions resulting in clear consensus, or to clear unbiased statements of law, why "20 is clearly to many," and avoid the use of original research and non-consensus point-of-view interpretations of policy. Thank you. DHowell (talk) 19:54, 22 April 2008 (UTC)