Wikipedia talk:Fancruft/Geogre's thoughts

Geogre's Thoughts on Lost and Found Information

I'm sorry if I duplicate anyone else's points, as I'm coming to this late, but it seems to me that there are a few ways of looking at this issue. I apologize in advance for going long on the subject, but I've thought about it rather a lot.

1. Why do we put in "fancruft?" The first way to think about this is strictly from the point of view of the contributor. I remember, when I was young, that a lot of conversations I had with friends ran along these lines: "Umm, 'member when they said Let's go to Camelot, and then one guy says, It's only a model? That was cool." I don't mean to denigrate with this characterization. There is an impulse of reliving by repeating that persists in all of our communications with friends. Part of the reason people put in fancruft articles is the same impulse. "Remember Dr. Chaos? Butters turned into Dr. Chaos. (That was so cool.)" However, another impulse is setting the record straight. "It is not 'excuse my while I kiss this guy!' Look at the article I just made, ok? It's very clearly, "Kiss the sky!" Or, for that matter, "The phase warp bubble is the crisis that beset the Enterprise on Star Trek the Next Generation in the episode 'Wesley Save the Ship.'" The impulse is, again, to set out a clear hierarchy of facts in a frequently discussed fiction. This is something I might call the bar-bet impulse. The reason a contributor wants to put it in is to codify something that came up in conversation.
2. What does a person get from fancruft? What does useful fancruft do? If there were doubt about the existence or priority of these items, then it would be worthwhile to have a record. So, if it was likely that someone was out there encountering "warp phase bubble" and had no idea what it meant and needed to know, then the fancruft article would answer the question. If someone heard about Queen Beryl and needed to have doubt dispelled, the article would serve that need.
3. What separates "fancruft" from an encyclopedic article. Let's consider Queen Beryl and my own Glumdalclitch. I consider one of them fancruft. Why? They're both about characters, after all. One is from an old book, and one is from a new show. The operative difference, to me, is that the term "Glumdalclitch" will be mentioned in a context wholly alien from its master work, and therefore someone reading the daily newspaper might well hit the term and be given absolutely no clue where it came from. That person needs an encyclopedia to explain the term. Is anyone going to hit "Queen Beryl" in any context outside of Sailor Moon and need the Sailor Moon meaning explained? I.e. can the function of #2 above actually be achieved? The "Hey, 'member when" impulse I disregard as being non-encyclopedic. To me, "fancruft" is detail from a fiction or mythos that does not exist in a context other than its fiction or mythos. Will you hit "Endymion" outside of reading Ovid? Yes. Will you hit "Borg" outside of watching Star Trek? Yes. Will you hit Digimon Clobbersaurus outside of the game/movie/TV show? No.
4. Games: Games are particularly evil in the fancruft realm. The reason is that there is yet another aspect of fancruft not mentioned above: games without documentation (whether real life or fictional) that generate multiple items get written up in an attempt at providing the documentation. That's what happens with all the Pokemen. The reason that I reject them is that the information is, like the stuff mentioned in #3, not known or desired except within the self-referential context of the game. Granted, the games have sprawling (and, in fact, open-ended by nature) realms. Imagine that we had someone writing an article on Mickey Mantle rookie year card. That's a baseball card, for the non-USonians. That card is traded in real life for hundreds of dollars. Two or three chewing gum companies issue baseball cards on every single player in the major leagues every year. Thus, there is the Mickey Mantle 2nd year card, 3rd year card, 4th year card, etc. But there is also one for Phil Nekro. There is one for Catfish Hunter. In fact, now that multiple gum companies are in the business, there may be 3 for every year. Fans/traders of baseball cards have enormous websites that do nothing but track the sale values of these cards. The money involved is enormous. The same industry exists for hockey and basketball in the US. I gather that it exists for football in the UK, too. It isn't that there is nothing to say about these cards, but rather that no one except a card trader -- i.e. not the general public, but devotees of a particular enterprise -- will need or want to know. We avoid getting into the super-precise details of gene splicing, sewage engineering, etc., because we're a general knowledge encyclopedia, and not a trade encyclopedia.
5. The balance of contributor and user: Do we care more about the people making articles or the people reading the encyclopedia? VfD is entirely biased, because the people who go there are, by necessity, authors. Therefore, its results are distorted by people who think like authors. Think about the users, rather than the contributors, and ask whether an article serves the researcher. How much should we bend to keep happy authors? How much should we be willing to tell people, "Write about what you want, but not that -- that's silly?" It's difficult and mean to do the latter, but I think we have to sometimes.
6. Lost information. My biggest crusade on VfD with "fancruft" is not to suppress information, but rather to ensure that information is logically organized. The static warp bubble might be important. I'm not a Star Trek fanatic, so I don't know how much of a cause celeb it is. However, having it as a stand alone article strands the information. Also, it occupies a name space position. Granted, there may not be another article dying for that spot, and there may be a disambiguation in the future, but let's be clear: the information is stranded. Someone has to be typing in that exact phrase to find it. If it's linked from the master fiction, then why isn't it just included in the master fiction? Do we have a choice only of every aspect an article and all aspects in one article? Can't we subdivide a topic a level or two before we get to the point where "The Entertainment Unit" is an article (instead of a thing in the Friends article)? Can't we have Items in Star Trek and Star Trek engines? Can't we have Sailor Moon villains? There must be a logical location where the information can be of use, per #2 above, for the people who need and want it, where they might in fact gain a context for the information that places it logically within the larger topic.
   

Anyway, thus run my thoughts on the subject, or at least the ones I could marshal at this late hour. Geogre 04:50, 13 Dec 2004 (UTC)

I only recently stumbled onto your user page, Geogre, and I wish I had done that much earlier. You not only have outspoken opinions on the subject, but are on of the most eloquent people I've seen on the Wiki (not to belittle the contributions of others, of course). I just wanted to say that, because I actually have no time to delve into your arguments now, which is a dirty rotten shame. I will do so eventually, however, because there's a lot of merit in them. JRM 09:27, 2004 Dec 13 (UTC)

Thanks a lot. I appreciate the compliment. The more I think about the case I made above, the more I think that it comes down to specialist vs. general reference. Folks who do basic research know that there are hundreds of real world items of great importance (to them) that are specialist information. When I worked in a lab, we did research on IGFBP-3 and IGF-I and how they could be used together to lower insulin resistance in type 2 diabetes mellitus. All real. We had 20+ publications on the subject. I've never been tempted to write it up. Never been tempted to write up Son of Sevenless (a protein), either. The reason is that this is information that the general researcher will not need and that these compounds have not yet passed into the general world. Mass-market fictions are different, but the principle is the same: a specialist needs to know the specialist information. The way we tell whether something is specialist or not is whether the thing appears in several contexts. If it doesn't, I still think we can talk about it -- just not at a stand-alone article. Anyway, that's one way to approach it. Geogre 13:30, 13 Dec 2004 (UTC)

JRM's response to Geogre

Yes, I get my own sections and subsections! But I'm not greedy; free sections for everyone!

Then I will subsub my responses, just because I can't hold a long thought in my head. Geogre 01:13, 14 Dec 2004 (UTC)

What (fan)cruft is

[...] will be mentioned in a context wholly alien from its master work, and therefore someone reading the daily newspaper might well hit the term and be given absolutely no clue where it came from. That person needs an encyclopedia to explain the term.

Up to a point, this is a compelling argument. However, this "it's all a matter of publicized degree" is of course as considerate as it is inconvenient. How about The Quatrain of Seven Steps, to take one of my pet examples? The article doesn't convince me it's of inherent significance out of context, but I trust the claim that it's "famed"—in ancient Chinese literature. It is featured in the Romance of Three Kingdoms, which is demonstrably significant in a broader cultural context. (Incidentally, the Quatrain is not put into context anywhere in the main article, which we should probably consider a flaw—it is only very indirectly referenced through Cao Zhi.) Now, how likely is it that someone will encounter it out of context, that is, "reading the daily newspaper" or something of similar scope? Is it even meaningful to have detailed discussions over whether this might or might not so? And encounter it out of context where? In English-speaking countries? Chinese? Among scholars of Chinese literature? Among American high school students? Wikipedians? Or "the general public", whoever they are? Should this be a basis for extensive discussion on whether something deserves its own article?

Volley

I call it a volley because I don't think that it's an answer. I don't necessarily think answers are going to be found, much less offered, but I will attempt to push what part of the birdie back over the net I think belongs there. (-10 points for mixed metaphor, I know.)
I may have been hasty with the "newspaper" example. Certainly, I don't expect Westminster Assembly to appear in the paper tomorrow. I wouldn't see Battle of Acteon in USA Today, either. My point was not such mass coverage, but rather the need for explanation. The need for explanation is going to have to arise from somewhere. The hypothetical I draw is someone reading something else -- even if it's a videogame based on Chinese history -- and needing to know something. However, what's missing from this criterion is that there is yet another function of an article, a function that doesn't come up in the fancruft debate and therefore wasn't mentioned in my thoughts: An article can explain what something is, and it can explain the relevance of a thing outside of itself. For example, an artilce on Newton's 2nd Law might just say, "Well, here it is. Newton thought it up." If it did that, it would nearly be a merge and redirect case (as most fancruft is). What it would do better is explain how the 2nd law can be seen in the work of rockets, how it is involved in orbiting satelites, how it explains drag, etc. I.e. it expands the knowledge into new contextual environments. Most fancruft can't really do that. I think that your Chinese quatrain could, though. It could have a literary and a historical context, and then it can have a cultural context of the production of poetry at that time, etc. In other words, it isn't just "what is this thing I hit in the paper over my coffee," but "what is there to learn from this thing?" Show me Queen Beryl that way, and I'll say that the interest is worth keeping solo, provided folks will look for it. Geogre 01:13, 14 Dec 2004 (UTC)

Whether Wikipedia can or should be specialist

At this point we are still not talking about whether fancruft should stay or go (to either oblivion or a more relevant context), merely what it is. We avoid getting into the super-precise details of gene splicing, sewage engineering, etc., because we're a general knowledge encyclopedia, and not a trade encyclopedia. Here, however, we hit upon a fundamental snag, namely the belief (I say belief) that Wikipedia is a general knowledge encyclopedia.

Now, before you declare me a nutter: of course items of general knowledge should be featured. If I want to know about mortgage or the Eiffel tower or General Custer, I should find comprehensive articles on them. But Wikipedia is not (to name the elephant) Britannica. It is not now, and will never be like Britannica. I do not say this based on Wikipedia's current biases and tendencies; I say this based on the fundamentally different medium and modus operandi.

One can argue that these do not matter. An encyclopedia is an encyclopedia, whether it is obtained by a committee of editors carefully scrutinizing the brilliant prose of experts in their fields, or a million monkeys on a million typewriters hammering out everything from patent nonsense to subtle hoaxes to the perfect article, building upon each other's work as they do so. (To drop a reference: METHINKSITISLIKEAWEASEL. :-) That's a possible stance, but it's unclear why we should limit ourselves in this manner.

I am talking, of course, about that great bugaboo of the standards advocates: Wiki is not paper. People have loved and hated this from day one.

On the one hand, it seems perfectly reasonable to argue that, because Wikipedia has no effective size limits, the threshold for topics is zero. Put in anything you like! We can easily store it! It'll become a factually neutral, NPOV article eventually, and that equates usefulness! Have faith!

On the other hand, Wikipedia calls itself an encyclopedia. No matter how you extend that word, it cannot possibly cover everything. Wikipedia is not a specially formatted World Wide Web. We have to ask ourselves what an encyclopedia is good for, and what would make Wikipedia meet those demands. There are a great deal of denials on what Wikipedia is not, but all this still leaves in a great deal in principle.

There's a good reason for that: there has not yet been any consensus as to what should and what should not be in, and consensus, as unfortunate as it may be, is the only reasonable way to go on something as big as Wikipedia.

The demands for factual accuracy and the neutral point of view have been remarkably uncontroversial (though not unchallenged), but most of the rest has not been. Wikipedia:Importance is still a hotbed of discussion. No consensus is to be expected in the near future, and it is not impossible that there may not be one without fundamentally respecifying Wikipedia's mission statement.

All this has created worrisome outcroppings. We now have the experimental wikicities.com, for example. This sounds like a very positive project, until you realize that it fundamentally overlaps with Wikipedia. Should there be a Wikipedia article on Pokémon? Of course. Should there be one on Wikicities? Well, obviously. But wait a minute—what goes here and what goes there? Are we going to have people who are experts on both their standards shuttle back and forth with information? That seems terribly wasteful. Wikicities is just a lollipop for people who have their Wikipedia article deleted, and a good way for Wikipedians to feel justified in their judgement (since the article will not be irrevocably deleted, just shuttled off to Siberia).

And where does it end? Should we have math Wikicity for specialized math knowledge? But what's specialized? Were adjoint functors ever featured in a newspaper? No—so transwiki, then! But what about the related context? Before we spawn a hundred forks and child projects that are not as uncontroversially different as Wiktionary or Wikiquotes (as if these were ever uncontroversial!) we might want to rethink the structure of the Wikisomethings as a whole. In particular, I believe most of these categorization schemes are too ad-hoc to properly scale.

More than one person has advocated to simply keep everything on Wikipedia. That's not realistic with the current structure of Wikipedia: we'd have quotes, dicdefs, semi-original research, trivia lists and Lord knows what mixed in with supposedly encyclopedic articles. We might as well redirect to Everything2 and throw in the towel.

Yes, we need standards. But I don't think we need to lay them down once and for all, or go off establishing holding cells (sorry, I mean "fertile grounds for enthusiast discussions") for material we summarily judge inappropriate. We need a structure that can establish them by consensus, and update as consensus moves. In short: we need new mechanisms for deciding what "stays", what "goes" and what should be "different" or "not available to the general public, but maybe to a specialized one", and VfD won't cut the mustard.

Volley

And here I disagree. I'm afraid that I have to take a "tough noogies" attitude, to some degree, with the loss to Wikicities. I know that's mean and heartless, and I wish I didn't feel that way. It may be too much time on VfD that makes me that way. (I have found that everyone should to go wp:fac and read the articles there at least once a day as an antidote to the poison one gets from VfD.)
I recognize that it is not spelled out that we are a general knowledge encyclopedia, but I think it's not spelled out in the way that air has no color. It's the licensing assumption behind every other policy. Were we not to have that assumption, then we would allow vanity articles, for example ("Hey, he's a real person!"), and CV's and endless music minutia. Furthermore, we have encyclopedia as a distinction to dictionary (move to Wiktionary), documentation (Wikisource), foreign language (the various wikis in other languages, including Latin and Klingon), etc. We have these other projects to hold distinctly non-encyclopedic content that is of worth. These places are only Siberia if contributors bring their own snow. There really shouldn't be a priority. This isn't the good one and those junk, I hope, or it doesn't have to be that way.
The reason that I say we're general knowledge is that there are worlds nad worlds of specialist information that we're not carrying, from business listings (to choose the vulgarity that does end up at VfD) to extraordinarily minute scientific terminology. We do have to debate whether something is too specialized, and I agree that no general rule will cover it, except to say that the consensus must be reached and that we ought not be so specialized as to be inaccessible or worthless. Tris is a really important buffer in cellular biology. If we have a Tris article, I hope it's not about the buffer. Chinese hamster ovaries are cells of choice for everyone doing in vitro cell biology, and they're always called CHO. I really hope we're not covering it. Some specialist information is on Wikipedia (literary stuff is represented even when it's nearly incomprehensible (and, from what I can tell, incomprehensible to the authors, too, because some of our articles, like semiotics, are just plain wrong (and no, I'm not going to fix them; I don't like edit wars with eager upperclassmen and master's students -- I had my fill to overflow))), but I think there is a level of detail and granularity that is too fine to be useful.
I will admit that I regard specialist information from pop culture more loathesome than specialist information from the sciences or literature or history, but I hope I'm consistent enough to vote to merge or delete regardless of the height of the culture that spawned it. Geogre 01:26, 14 Dec 2004 (UTC)

What's wrong with our current mechanisms

When you condense it into one phrase, "'"fancruft" is detail from a fiction or mythos that does not exist in a context other than its fiction or mythos" you are presenting a very valuable criterion. Unfortunately, also a very debatable one. And one which Votes for Deletion may be manifestly unsuitable to decide upon.

To take the aforementioned Quatrain, how would non-experts establish that it exists in an independent context? (We are assuming that the article mentions its context, of course, otherwise things look even more gloomy.) If we are lucky, Google will help us, or an expert Wikipedian (perhaps the original contributor) is present. But what if we're not lucky, at least not immediately? Should we be bold and put the Quatrain in in the main article? If there's no supporting frame for it in the article, should we just go ahead and insert it anyway, decreasing the quality? The answer is probably "yes, because excessively detailed information in context is more valuable than excessively detailed information in an unlinked article".

There must be a logical location where the information can be of use, per #2 above, for the people who need and want it, where they might in fact gain a context for the information that places it logically within the larger topic.

We are in perfect agreement here. I'd like to see much more aggressive merging and redirecting. VfD is demonstrably broken when it comes to handling these cases. Instead of finding a context with which to merge, the first instinct of people will be to list an out-of-context article for deletion. This would be fine if they kept to voting merge and redirect once a context had been established—but they do not. Even if a context is manifestly available, people will vote to delete, possibly on the grounds mentioned above, possibly on personal predilections. (I am not arguing that many keep votes are not similarly biased, of course.)

This is demonstrably counterproductive—we lose the information, and force potential future contributors to reinvent the wheel (not necessarily as well as the first attempt). It's also an incredible waste of time and resources for VfD voters, because if such merging is used as the first option, the result is simply subject to the wiki editing process, which has a strong tendency to keep information unless manifestly wrong or out of context. (Discussing ways to encourage the tendency to weed out "cruft" from articles, fan or no, is another matter.) Finally, having any resulting inappropriate redirects deleted (of which there only be few) is much less controversial, time-consuming and atmosphere-poisoning than deleting the article.

Do we care more about the people making articles or the people reading the encyclopedia? VfD is entirely biased, because the people who go there are, by necessity, authors. Therefore, its results are distorted by people who think like authors.

This is a very good point. Eventually, we should probably have a mechanism that works by appreciation of the readers, not the writers. Many writers are competent and unbiased judges of work, based on how they would decide as a reader. Unfortunately, many more writers aren't. I have yet to see the first Wikipedian unencumbered by any sort of pride or ego, and such a Wikipedian would probably not be the very dedicated sort in the first place.

What these mechanisms should be and how we should transition to them is, fortunately, out of the scope of this discussion. (I say fortunately, because I've already talked for too long. Fortunately, Wiki is not paper... JRM 00:06, 2004 Dec 14 (UTC)

Volley and Confessio

I'm going to be honest: You will very often see a vote from me that reads, "Delete: Probably could be discussed in the Dragonball Z article." I have my moments of deep stupidity like anyone, but most of the time I am very aware of what I have just voted. Most of the time I am saying, and I admit it, "If the author wants to merge the material and create a redirect, groovy, but let's not put yet another overly microscopic article that was too hastily written into the merge queue." This is because of the back side of VfD -- the dispensation phase. On VfD/old, we have a backlog like you wouldn't believe. The backlog for keeping and deleting is over a month. The merge load is unbelievable. I admit that I am making a value judgment on the article author. I hope I'm wrong, but I keep thinking of a kid who is desperately looking for a topic that isn't covered on Wikipedia so that he or she can write an article. That kid grabs their favorite thing -- some trading card game or cartoon -- and writes about The Castle Where It Blowed Up or something. When the VfD tag hits the page, the author doesn't debate it, doesn't show up at all -- just wanted to write an article. I feel like the rest of the community is being told to walk along the highway with a stick with a nail on the end of it. Worse, we're being asked to spear the garbage and not throw it away, but, rather, to recycle it and make a lovely cottage out of it.
Maybe it's a bad attitude on my part, but I feel like merging should be done only with truly valuable information, and I am, I admit, judging the information. How many readers will we lose if we don't explain that Lord Viperscorpion has a web comic? How many readers will we lose if we can't explain what power crystal is needed to turn Clobbersaurus into Supermegatroid? If I don't think we're going to hurt our users much, I won't vote merge, even when I think that a merge would be acceptable. If there is a volunteer, I'm delighted.
I used to vote "Clean Up" a lot more. I also used to work on the Clean Up pages more. I don't feel guilty about not doing that anymore, either, as I've gotten into creating more new content and working on Featured Articles. One reason I quit was that a lot of folks were sending desperately bad stuff to clean up because they were opposed to using the speedy delete tag. Clean Up was getting 20 miles long. Thus, people doing clean up pretty much stopped actually trying to save articles and began...wikifying. (sigh) I can't say "Send to clean up" if all it needs is wikifying. If that's all an article needs, I'll do it myself.
So, we don't have a way to channel people (especially the blanket "keep" voters of recent vintage) to clean up or the merge queue. That makes me, anyway, a lot meaner about these things. I don't want to be, but we need a lot more help, and I mean a lot more help on serious clean up and merging, and we need a way for the Clean Up page to just plain ditch stuff. (I have just today gotten tasked with "You should spend you energy helping the project instead of deleting on VfD." This came from someone who refused in the past to expand a stub after voting "keep" and from someone who obviously hasn't seen my user page.) The older "keep" voters ("inclusionists") used to never do that until they'd edited the article to try to save it.
Sorry.... got off the beam there. Enough cane shaking from me. Geogre 01:39, 14 Dec 2004 (UTC)

JRM's final note

I could reply to the above points, but I actually feel they make nice summaries of both our stances, and equally nice illustrations of some of the fundamental disagreements. I will not say "inclusionist" versus "deletionist", because I feel that vastly oversimplifies the issues—I am not principally for including everything debatable, and Geogre is not principally for deleting everything debatable. I rather think that both groups tend to emphasize different aspects of Wikipedia, and that is what's causing all the discrepancies. In any case, as if it's not already clear, the last word certainly has not been said about these matters. But I trust you can make up your own mind on that. JRM 11:40, 2004 Dec 14 (UTC)