Just a Heads Up slashdot has new article about wikipedia up and it's use of experts - "The Role of Experts In Wikipedia" http://tech.slashdot.org/article.pl?sid=09/02/16/0210251
K. Peachey wrote:
Just a Heads Up slashdot has new article about wikipedia up and it's use of experts - "The Role of Experts In Wikipedia" http://tech.slashdot.org/article.pl?sid=09/02/16/0210251
"Sanger was one of the founders of Wikipedia, and of its failed predecessor Nupedia, who left the fold because of differences over the question of the proper role of experts."
Strange, I thought it was because he stopped being paid for it.
http://meta.wikimedia.org/wiki/My_resignation--Larry_Sanger
"As you know, since the beginning of February, I've been working on Wikipedia and Nupedia as a part-time volunteer. I haven't been able to do nearly as much as I wish I could do, but job-hunting and money-making activities necessarily occupy a great deal of my time. Unfortunately, I do not expect to see, within the foreseeable future, any sort of compensation for the time and responsibility I've continued to hold in the projects. Now that I'm unemployed, I can ill afford to spend my free time this way. This is, I'm afraid, by far the most important reason for my resignation."
He then forgot about Wikipedia completely for a few years, and re-emerged a critic once the media started paying attention to it.
-- Tim Starling
Regardless of the history, Sanger does have a viewpoint that would be worth reading even if the author were anonymous. In particular, the following claim is quite accurate to my experience:
Over the long term, the quality of a given Wikipedia article will do a random walk around the highest level of quality permitted by the most persistent and aggressive people who follow an article.
- Carl
Carl Beckhorn wrote:
Regardless of the history, Sanger does have a viewpoint that would be worth reading even if the author were anonymous. In particular, the following claim is quite accurate to my experience:
Over the long term, the quality of a given Wikipedia article will do a random walk around the highest level of quality permitted by the most persistent and aggressive people who follow an article.
It is a nice use of rhetoric, but accurate? NOWAI!
Let me paraphrase it in a way that will make the logical flaws more apparent.
In that sentence there are buried assumptions as follows:
1. There are people on wikipedia who will not permit quality.
2. People who won't permit quality are aggressive.
3. There is a clear unambiguous metric for quality.
4. Aggressive people who won't permit quality will follow an article.
5. Over the long term, the dynamics of wikipedias practices will not prevent editors who will not allow quality on wikipedia from dragging it down to the level that they aggressively and persistently insist on bringing it down to. There are no working heuristics to allow it to transcend that attractor.
*Understanding* the logical flaws of those 5 statements is left to the student.
Yours,
Jussi-Ville Heiskanen
Well... that does happen. It's basically WP:OWNership. I find that ownership usually, but not always, stops an article reaching its maximum quality and/or coverage.
But ownership doesn't seem to dominate the wikipedia. And sometimes if the owner really is really good then the article can end up just fine.
On 16/02/2009, Jussi-Ville Heiskanen cimonavaro@gmail.com wrote:
Carl Beckhorn wrote:
Regardless of the history, Sanger does have a viewpoint that would be worth reading even if the author were anonymous. In particular, the following claim is quite accurate to my experience:
Over the long term, the quality of a given Wikipedia article will do a random walk around the highest level of quality permitted by the most persistent and aggressive people who follow an article.
It is a nice use of rhetoric, but accurate? NOWAI!
Let me paraphrase it in a way that will make the logical flaws more apparent.
In that sentence there are buried assumptions as follows:
- There are people on wikipedia who will not permit
quality.
People who won't permit quality are aggressive.
There is a clear unambiguous metric for quality.
Aggressive people who won't permit quality will
follow an article.
- Over the long term, the dynamics of wikipedias
practices will not prevent editors who will not allow quality on wikipedia from dragging it down to the level that they aggressively and persistently insist on bringing it down to. There are no working heuristics to allow it to transcend that attractor.
*Understanding* the logical flaws of those 5 statements is left to the student.
Yours,
Jussi-Ville Heiskanen
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
K. Peachey wrote:
Just a Heads Up slashdot has new article about wikipedia up and it's use of experts - "The Role of Experts In Wikipedia" http://tech.slashdot.org/article.pl?sid=09/02/16/0210251
Sanger says the main reason that Wikipedia's articles are as good as
they are is that they are edited by knowledgeable people to whom deference is paid, although voluntarily, but that some articles suffer precisely because there are so many aggressive people who 'guard' articles and drive off others (PDF), including people more expert than they are.
The good articles are good basically because smart people take the trouble to research them and write them to a decent standard. The article on topic X is good, when it is, not usually because A, an expert on X, has filled it with A's expert knowledge, but because B and C and maybe others have looked at some literature on the topic and done a decent job of constructing a precis for the general reader. I would make an exception for some areas (e.g. mathematics, medicine) where an expert is going to have a view that is 1000% clearer than someone coming in from outside. The bit about "deference" shows a fixation on the more combative aspects of WP. Most articles aren't that contentious.
'Without granting experts any authority to overrule such people, there
is no reason to think that Wikipedia'a articles are on a vector toward continual improvement,' writes Sanger.
No reason for Sanger to think that, since he continually misses the point of the wiki. Most articles, numerically speaking, just wait until someone who cares comes along and upgrades them.
Wikipedia's success cannot be explained by its radical egalitarianism
or its rejection of expert involvement, but instead by its freedom, openness, and bottom-up management and there is no doubt that many experts would, if left to their own devices, dismantle the openness that drives the success of Wikipedia.
Yeah, we know about such experts, but they are not experts _on Wikipedia_! How about a little respect for the expertise of people who spend time doing it, rather than talking about it?
'But the failure to take seriously the suggestion of any role of
experts can only be considered a failure of imagination,' writes Sanger. 'One need only ask what an open, bottom-up system with a role for expert decision-making would be like.'
In other words, despite all appearances, CZ is superior to WP. Well, I think we saw where this was going a little earlier.
The brass neck involved in implying that WP is "unimaginative", which is largely wrong, rather than too utopian, which is certainly an arguable point, is breath-taking. A propos FR, or other such things, there has been this constant debate in which the "pure wiki" model is held up against what amount to pragmatic suggestions for change in aid of the encyclopedic mission. This discussion goes on all the time.
Charles
2009/2/16 Charles Matthews charles.r.matthews@ntlworld.com:
K. Peachey wrote:
'But the failure to take seriously the suggestion of any role of
experts can only be considered a failure of imagination,' writes Sanger. 'One need only ask what an open, bottom-up system with a role for expert decision-making would be like.'
In other words, despite all appearances, CZ is superior to WP. Well, I think we saw where this was going a little earlier.
Which in practice will end up a bit like this:
http://reinderdijkhuis.com/wordpress/2009/02/12/citizendium-the-encyclopedia...
Precis: experts are not a panacea.
cf: Stirling Newberry's many posts several years ago to wikien-l and wikipedia-l pointing out that the problem with a lot of experts is that they got to be experts by pushing a POV better than anyone else.
(Larry Sanger is aware of this blog post, and dismisses it as missing the point. However, the Citizendium article on homeopathy is still an NPOV disaster.)
In any case, we're not short of experts on Wikipedia. You can hardly move without bumping into a Ph.D. It's not nicknamed "Unemployed Ph.D Deathmatch" for nothing.
- d.
Jussi-Ville Heiskanen wrote:
In that sentence there are buried assumptions as follows:
- There are people on wikipedia who will not permit
quality.
People who won't permit quality are aggressive.
There is a clear unambiguous metric for quality.
Aggressive people who won't permit quality will
follow an article.
- Over the long term, the dynamics of wikipedias
practices will not prevent editors who will not allow quality on wikipedia from dragging it down to the level that they aggressively and persistently insist on bringing it down to. There are no working heuristics to allow it to transcend that attractor.
*Understanding* the logical flaws of those 5 statements is left to the student.
It would be rash to say you couldn't find any examples where this is true - there is a large selection of articles. It might be a fair model for the article about, for example, a controversial Governor of Alaska who didn't get chosen as a candidate for Vice-President. But you could click Random Article for a little while before you came up with an article to which this argument really would apply.
Charles
On Mon, Feb 16, 2009 at 3:29 PM, David Gerard dgerard@gmail.com wrote:
<snip>
However, the Citizendium article on homeopathy is still an NPOV disaster.
I hadn't visited Citizendium for ages.
It is an interesting exercise to read through the Wikipedia article on Homeopathy and the Citizendium one, and see the strengths and weaknesses of both, and also the similarities and places where the differences are very subtle (different order of sections) and less subtle ("Professional homeopaths: who are they?"). But equally, the Wikipedia article lacks a section on homeopaths, professional or otherwise.
One thing that strikes me is that both articles are difficult to read and poorly written. In other words, when something is controversial and has a high rate of editing, the readability quality invariably decreases in the ensuing chaos.
The name of the three editors who have approved the Citizendium article makes interesting reading as well:
Gareth Leng D. Matt Innis Dana Ullman
Weirdly, most of the history is not there:
http://en.citizendium.org/wiki?title=Homeopathy&action=history
But has been moved to a draft page:
http://en.citizendium.org/wiki?title=Homeopathy/Draft&action=history
So maybe I should have read that draft instead. It would be nice to know which versions were approved by the three editors above, and at what stage.
Carcharoth
2009/2/16 Carcharoth carcharothwp@googlemail.com:
One thing that strikes me is that both articles are difficult to read and poorly written. In other words, when something is controversial and has a high rate of editing, the readability quality invariably decreases in the ensuing chaos.
That doesn't just apply to controversial subjects. Any subject with a high edit rate shows the same problem - current events, for example (they generally end up being timelines in very poor disguise). What it needs is for someone to periodically go through the whole thing tidying it up - not a fun job, admittedly, but a necessary one. (This does happen with some articles, and you can tell the difference.)
On 16/02/2009, Charles Matthews charles.r.matthews@ntlworld.com wrote:
The good articles are good basically because smart people take the trouble to research them and write them to a decent standard. The article on topic X is good, when it is, not usually because A, an expert on X, has filled it with A's expert knowledge, but because B and C and maybe others have looked at some literature on the topic and done a decent job of constructing a precis for the general reader.
To be slightly more accurate, the wikipedia does indeed depend heavily on experts and smart and knowledgeable people, but only after their material has been published; and only then if the publisher is judged to have good filtering processes in place to minimise the non negligible chance that these people are wrong or unbalanced about what they say.
Looked at like this, Sanger wants to take this a stage further- he wants to actually *weaken* the fact checking by using unfiltered experts saying more or less whatever they want. I would expect that this can result in poorer articles than the wikipedias model, in addition to the many obvious problems about how you find and validate and keep these experts engaged.
Charles
What Citizendium's Homeopathy article shows more than anything is that a wide base of editors, and therefore a wide audience, is essential for the success of Wikipedia or any similar project. The article shows a distinct lack of the cleansing effects of sunlight; few people read it, few people contribute to it, and it's become the home of fringe POV editors who have tried and failed to sling the same on Wikipedia. Citizendium might do better if it made a more concerted attempt to discern between actual experts in fact and self-described experts in name only, but they apparently have chosen not to do that or don't have enough people from which to pick.
I say, let them congregate on Citizendium. We should have a template {{trycitizendium}} that we can post on the pages of our more aggressive POV pushers.
Nathan
David Gerard wrote:
Which in practice will end up a bit like this: http://reinderdijkhuis.com/wordpress/2009/02/12/citizendium-the-encyclopedia...
Precis: experts are not a panacea.
Mmmm, Larry of course does have a valid point in there, which is that edit warring correlates with a lower quality of article. It looks like his solution is this: decide who should win the edit wars in advance. Bingo! No serious edit wars. Sadly, as this case apparently indicates, the syllogism that this will get you a higher quality of article isn't valid. Guess what - sometimes you have to put up with the pesky business of people needing to argue the matter out on talk pages.
Charles
Nathan wrote:
I say, let them congregate on Citizendium. We should have a template {{trycitizendium}} that we can post on the pages of our more aggressive POV pushers.
The template need not limit itself to Citizendium, though the symbolism of having it in the template name has a certain value. If the posted list contains several other such approved projects, with a one-line blurb about what they are, the aggressive person may more easily and quietly find one to his liking without actually stumbling upon WikipediaReview.
Ec
On Mon, Feb 16, 2009 at 5:40 PM, Ray Saintonge saintonge@telus.net wrote:
Nathan wrote:
I say, let them congregate on Citizendium. We should have a template {{trycitizendium}} that we can post on the pages of our more aggressive POV pushers.
The template need not limit itself to Citizendium, though the symbolism of having it in the template name has a certain value. If the posted list contains several other such approved projects, with a one-line blurb about what they are, the aggressive person may more easily and quietly find one to his liking without actually stumbling upon WikipediaReview.
It can include sister WMF projects as well. Though care should be taken to not push malcontents on unsuspecting people. It is best to have a mentor or guide that can introduce you. It all depends whether the reason for the breakdown in editing relations is due to the person, the topic, or the environment. Sometimes it is all three.
Carcharoth
Charles Matthews schreef:
Guess what - sometimes you have to put up with the pesky business of people needing to argue the matter out on talk pages.
I've been following CZ for some time, and one gets the feeling that Larry Sanger doesn't really like arguementsi, or open discussion.
One of the rules at CZ is that you cannot complain about another editor at all. This is to prevent long discussions on who is right, but according to Adam Cuerden's blog post, it is one of the reasons why the Homeopathy article is so bad.
The "anti-homeopathy" side started to make complaints about the other side, and regardless of whether the complaints were justified, they were templated with {{nocomplaints}}. (No [[WP:DTTR]] at CZ...) And then the other side won.
The idea behind the nocomplaints policy seems a good one, perhaps, but in reality, it hampers discussion. After one of the latest intervention by Larry Sanger, the reply was "I don't think you should have removed that", which was promptly replaced by another {{nocomplaints}}...
I think there is a lesson in here for the other thread in this mailing list at the moment: A zero-tolerance policy on incivility will be interpreted loosely at times, which will hamper discussion, which leads to bad articles if the number of editors is low (like on CZ or at low-visibility WP articles).
Eugene
2009/2/16 Eugene van der Pijll eugene@vanderpijll.nl:
Charles Matthews schreef:
Guess what - sometimes you have to put up with the pesky business of people needing to argue the matter out on talk pages.
I've been following CZ for some time, and one gets the feeling that Larry Sanger doesn't really like arguementsi, or open discussion.
I've been following the CZ statistics page for some time, and I get the feeling that it doesn't matter because activity on CZ is shrinking (even Sanger doesn't seem very active) and it will never reach a size where anyone actually uses it. It's a fraction of the size of Wikipedia at the same age (in terms of articles or total words) and growth is slowing (whereas Wikipedia showed exponential growth at that time). We could discuss why it failed but I think the real answer is simply that Wikipedia is "good enough" so there is very little interest in a new project doing the same thing (and which won't be anywhere near as useful for several years, even with the more generous assumptions).
Carcharoth schreef:
Weirdly, most of the history is not there:
http://en.citizendium.org/wiki?title=Homeopathy&action=history
But has been moved to a draft page:
http://en.citizendium.org/wiki?title=Homeopathy/Draft&action=history
That's how they do that there. The approved page is a copy; the draft is moved to /Draft, and is a living document.
So maybe I should have read that draft instead. It would be nice to know which versions were approved by the three editors above, and at what stage.
The exact version that is now at [[Homeopathy]] was approved by the three editors. (These are "editors" in the CZ sense: experts in the article's workgroup, Health.)
It's a bit suprising to see one of the parties to the "edit war" as an approving editor...
Eugene
On Mon, Feb 16, 2009 at 2:10 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
I've been following the CZ statistics page for some time, and I get the feeling that it doesn't matter because activity on CZ is shrinking (even Sanger doesn't seem very active) and it will never reach a size where anyone actually uses it. It's a fraction of the size of Wikipedia at the same age (in terms of articles or total words) and growth is slowing (whereas Wikipedia showed exponential growth at that time). We could discuss why it failed but I think the real answer is simply that Wikipedia is "good enough" so there is very little interest in a new project doing the same thing (and which won't be anywhere near as useful for several years, even with the more generous assumptions).
I also have been following their stats page (as well as their forums), and I agree with the fundamental issue: Wikipedia had the first mover advantage, and has proved good enough for enough people to prevent any serious competition.
However, I don't think we should think of Citizendium as having failed. Certainly, it has failed to realize Sanger's and a few others' hopes to be on its way to eclipsing Wikipedia. But CZ has a fairly stable community; it's shrinking a little, but so is Wikipedia's community. It's a free content project that is producing some useful material, and some editors find it a nicer place to work than Wikipedia. It's licensed CC-by-SA 3.0, which means it will be compatible with Wikipedia soon. And a rather high proportion of content is stuff that isn't present on Wikipedia. The anti-Wikipedia ethos of the project has also waned as they've begun to sort out their own identity beyond "Wikipedia with real names where experts have power".
Sanger keeps claiming that they aren't growing simply because they haven't yet gotten serious about recruitment. I don't find that convincing, but it's not inconceivable that concerted efforts at recruitment could result in another wave of growth or two (though probably never exponential growth).
-Sage (User:Ragesoss)
Thomas Dalton schreef:
I've been following the CZ statistics page for some time, and I get the feeling that it doesn't matter because activity on CZ is shrinking (even Sanger doesn't seem very active) and it will never reach a size where anyone actually uses it.
I've had a bit of an argument with him recently about the decline of CZ (http://blog.citizendium.org/2009/02/04/write-a-thon-is-on-independent-notice...) and he assured me that CZ has been growing exponentially, and will be growing explosively, and that [[CZ:Statistics]] proved that.
It's a fraction of the size of Wikipedia at the same age (in terms of articles or total words) and growth is slowing (whereas Wikipedia showed exponential growth at that time).
CZ is actually only about half WP's size at the same age, I think. I've plotted the growth of both sites in number of words, and it is a surpising difference. CZ started much larger than WP because they imported a lot of WP articles[*], and then grew linearly. After a year, both encyclopedias were the same size, but because of WP's exponential growth, it is now outpacing CZ.
CZ's growth in number of words has only just begun to fall; its lack of new authors has been a problem for a much larger time.
We could discuss why it failed but I think the real answer is simply that Wikipedia is "good enough" so there is very little interest in a new project doing the same thing (and which won't be anywhere near as useful for several years, even with the more generous assumptions).
Could be. To succeed, a new encyclopedia will have to either have a very dedicated team of authors, or find a specialistic niche (scholarpedia?), or be useful from the start; perhaps by starting off with WP's entire content. And I don't think we have seen WP's successor yet.
Eugene
[*] In fact, they imported all of Wikipedia. But I'm only counting the articles they changed.
2009/2/16 Eugene van der Pijll eugene@vanderpijll.nl:
Thomas Dalton schreef:
I've been following the CZ statistics page for some time, and I get the feeling that it doesn't matter because activity on CZ is shrinking (even Sanger doesn't seem very active) and it will never reach a size where anyone actually uses it.
I've had a bit of an argument with him recently about the decline of CZ (http://blog.citizendium.org/2009/02/04/write-a-thon-is-on-independent-notice...) and he assured me that CZ has been growing exponentially, and will be growing explosively, and that [[CZ:Statistics]] proved that.
I don't see a claim of exponential growth (which would be complete rubbish), just "good news". I don't think linear growth (even slightly below linear) is good news, personally.
It's a fraction of the size of Wikipedia at the same age (in terms of articles or total words) and growth is slowing (whereas Wikipedia showed exponential growth at that time).
CZ is actually only about half WP's size at the same age, I think. I've plotted the growth of both sites in number of words, and it is a surpising difference. CZ started much larger than WP because they imported a lot of WP articles[*], and then grew linearly. After a year, both encyclopedias were the same size, but because of WP's exponential growth, it is now outpacing CZ.
CZ's growth in number of words has only just begun to fall; its lack of new authors has been a problem for a much larger time.
My calculations come out as about 1/10 the size by articles and 1/3 the size by words (so their articles must be longer on average). It doesn't really matter, though, when you have exponential vs linear, the exponential is always going to win regardless of precise numbers. (Of course, we're not growing exponentially any more, so I guess it's possible they could eventually catch up, but we're talking decades...)
We could discuss why it failed but I think the real answer is simply that Wikipedia is "good enough" so there is very little interest in a new project doing the same thing (and which won't be anywhere near as useful for several years, even with the more generous assumptions).
Could be. To succeed, a new encyclopedia will have to either have a very dedicated team of authors, or find a specialistic niche (scholarpedia?), or be useful from the start; perhaps by starting off with WP's entire content. And I don't think we have seen WP's successor yet.
A dedicated team of authors won't do it - Wikipedia grew exponentially because as it got bigger more people read it and more readers became writers. If all the writing is done by a set team the growth will only even be linear. A specialist niche is another matter entirely - wikis are good for all kinds of things, it's only the market for wiki general encyclopaedias that is filled.
2009/2/16 Sage Ross ragesoss+wikipedia@gmail.com:
However, I don't think we should think of Citizendium as having failed. Certainly, it has failed to realize Sanger's and a few others' hopes to be on its way to eclipsing Wikipedia. But CZ has a fairly stable community; it's shrinking a little, but so is Wikipedia's community.
We're shrinking because we've already written most of the stuff we want to include. We're over the hill and rolling down the other side, they never got up the hill and are rolling back down to the start.
It's a free content project that is producing some useful material, and some editors find it a nicer place to work than Wikipedia. It's licensed CC-by-SA 3.0, which means it will be compatible with Wikipedia soon. And a rather high proportion of content is stuff that isn't present on Wikipedia. The anti-Wikipedia ethos of the project has also waned as they've begun to sort out their own identity beyond "Wikipedia with real names where experts have power".
Sanger's anti-Wikipedia attitude doesn't seem to have changed much. I don't know about the rest of the community, I don't read their stuff in much detail.
Sanger keeps claiming that they aren't growing simply because they haven't yet gotten serious about recruitment. I don't find that convincing, but it's not inconceivable that concerted efforts at recruitment could result in another wave of growth or two (though probably never exponential growth).
They've been going for over two years, if they were going to have a big recruitment push wouldn't they have done so by now? But really, trying to recruit writers is the wrong way round, they need to recruit readers, that's where the writers come from for exponential growth (which they need if they are going to get anywhere). However, I can't see how they can recruit readers until they have enough articles to be useful - it's a catch-22 and that's why I don't think any similar project will ever rival Wikipedia, simply because we got there first.
Thomas Dalton schreef:
I don't see a claim of exponential growth (which would be complete rubbish), just "good news". I don't think linear growth (even slightly below linear) is good news, personally.
I exaggerated somewhat. But he has spoken about ongoing exponential growth before, so it annoyed me a bit to see him in denial again.
My calculations come out as about 1/10 the size by articles and 1/3 the size by words (so their articles must be longer on average).
About 30% of the volume of WP at the time consisted of Rambot articles, which aren't too interesting as a measurement of growth (though they may have been a significant reason for our success).
There are two kinds of CZ articles: copied from WP (mostly the best articles; generally large), and original ones. The percentage original articles has been rising steadily. Judging from clicking "random page" a few times, they are not too different in length from the average WP article, so the average length of the CZ article is falling.
Eugene
On Mon, Feb 16, 2009 at 2:42 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
They've been going for over two years, if they were going to have a big recruitment push wouldn't they have done so by now? But really, trying to recruit writers is the wrong way round, they need to recruit readers, that's where the writers come from for exponential growth (which they need if they are going to get anywhere). However, I can't see how they can recruit readers until they have enough articles to be useful - it's a catch-22 and that's why I don't think any similar project will ever rival Wikipedia, simply because we got there first.
I don't disagree. I'm just saying we should think of Citizendium as another (small) place for people to produce free content similar to the kind Wikipedia produces, as a potential collaborator with Wikipedia rather than a competitor (which isn't realistic, if it ever was). That's a very real possibility once the license change happens.
-Sage (User:Ragesoss)
On Mon, Feb 16, 2009 at 11:42 AM, Thomas Dalton thomas.dalton@gmail.comwrote:
We're shrinking because we've already written most of the stuff we want to include.
This is orthogonal to the main conversation here, but this is not nearly the case.
We've picked off a lot of low hanging fruit, approaching all of it. Things which haven't been dealt with include figuring out where a more liberal livable line lies in the inclusionism question, and filling in a lot of semi specialist topics which are currently woefully underrepresented. There are whole fields of engineering and science that we have barely scratched the surface of at the moment.
A month-ish ago, I spent a week putting together an article on one explosives engineering topic which was completely missing... leaving us with about 95% of that field still uncovered so far.
Aerospace engineering is poorly covered.
Automobile engineering is somewhat covered, but not with really good articles. Rocketry needs a lot more. Astrodynamics needs a lot more.
Naval architecture and ship design topics are poorly covered now.
Three of the last eight highly technical terms I went looking for information on weren't in Wikipedia in any significant way, across a bunch of fields.
This is just what's on my mind right now. Every time I've looked at it I've found more gaps.
I could spend the rest of my life adding information to Wikipedia, at this rate, if I didn't have to have a day job and didn't want to go sit on a beach. Hopefully we can over time add more new editors / contributors in these fields so I don't have to 8-P
2009/2/16 George Herbert george.herbert@gmail.com:
We've picked off a lot of low hanging fruit, approaching all of it. Things
We've picked up all the fruit that's actually on the ground with neon signs pointing to it. There's lots of low hanging fruit, e.g.:
A month-ish ago, I spent a week putting together an article on one explosives engineering topic which was completely missing... leaving us with about 95% of that field still uncovered so far. Aerospace engineering is poorly covered. Automobile engineering is somewhat covered, but not with really good articles. Rocketry needs a lot more. Astrodynamics needs a lot more. Naval architecture and ship design topics are poorly covered now. Three of the last eight highly technical terms I went looking for information on weren't in Wikipedia in any significant way, across a bunch of fields.
Yep :-)
This is just what's on my mind right now. Every time I've looked at it I've found more gaps.
Here's to red links, the signposts to future growth!
I could spend the rest of my life adding information to Wikipedia, at this rate, if I didn't have to have a day job and didn't want to go sit on a beach. Hopefully we can over time add more new editors / contributors in these fields so I don't have to 8-P
Careful, you might get a life with that sort of attitude!
- d.
2009/2/16 Eugene van der Pijll eugene@vanderpijll.nl:
My calculations come out as about 1/10 the size by articles and 1/3 the size by words (so their articles must be longer on average).
About 30% of the volume of WP at the time consisted of Rambot articles, which aren't too interesting as a measurement of growth (though they may have been a significant reason for our success).
30% by articles, maybe, but they were stubs weren't they, so it won't be 30% by words. (That may explain why their articles are longer on average.) Incidentally, I don't think Rambot articles were that significant - if you look at the graphs, rate of growth didn't increase when they were added as one would expect is rate of growth were simply proportional to size (which it what gives exponential growth) which suggests rate of growth was actually proportional to the number of non-Rambot articles.
George Herbert wrote:
There are whole fields of engineering and science that we have barely scratched the surface of at the moment.
I think that's right. Engineering is not one of Wikipedia's strong areas, I believe, though I hardly spend time on that.
I do spend time on history - looks like 2009 will be the year of the seventeenth century - and there is clearly a great deal of "placeholder text": stuff that will do until someone gets round to putting in an effort, but easy to add to in detail. I'm working today on [[John Wilkins]], the really big single name in getting the Royal Society founded - and there was a huge amount to do just to make the article reasonably comprehensive.
I believe we have another decade before Wikipedia lives up to its potential as a comprehensive reference. My main hope is that life around the wiki stays dull enough so that the job largely gets done.
Charles
2009/2/16 Charles Matthews charles.r.matthews@ntlworld.com:
I believe we have another decade before Wikipedia lives up to its potential as a comprehensive reference. My main hope is that life around the wiki stays dull enough so that the job largely gets done.
Indeed. Current predictions show growth in terms of article numbers pretty much ending in around 4 or 5 years time. We'll then need several more years to actually get all the articles up the scratch. A decade may even be optimistic.
2009/2/16 Thomas Dalton thomas.dalton@gmail.com:
30% by articles, maybe, but they were stubs weren't they, so it won't be 30% by words. (That may explain why their articles are longer on average.) Incidentally, I don't think Rambot articles were that significant - if you look at the graphs, rate of growth didn't increase when they were added as one would expect is rate of growth were simply proportional to size (which it what gives exponential growth) which suggests rate of growth was actually proportional to the number of non-Rambot articles.
I remember them being a PITA at the time (early 2004). 200k articles with 30k Rambot articles meant [[Special:Random]] turned up Rambot articles entirely too often for my liking. I'm glad that (a) they're now vastly outnumbered (b) almost all have been significantly rewritten.
- d.
Sage Ross wrote:
I don't disagree. I'm just saying we should think of Citizendium as another (small) place for people to produce free content similar to the kind Wikipedia produces, as a potential collaborator with Wikipedia rather than a competitor (which isn't realistic, if it ever was). That's a very real possibility once the license change happens.
That's quite OK: someone who forks WP to make a site that is similar but has a different atmosphere doesn't have to prove a big philosophical difference, just to do it (which technically can't be so hard, these days). In other words, it would be nice if all the free-culture people agreed that this is not a zero-sum game. I don't read Sanger's comments in exactly that way, though.
Charles
Thomas Dalton wrote:
2009/2/16 Charles Matthews charles.r.matthews@ntlworld.com:
I believe we have another decade before Wikipedia lives up to its potential as a comprehensive reference. My main hope is that life around the wiki stays dull enough so that the job largely gets done.
Indeed. Current predictions show growth in terms of article numbers pretty much ending in around 4 or 5 years time. We'll then need several more years to actually get all the articles up the scratch. A decade may even be optimistic.
Yeah, well, my reaction to the whole "fruit" discussion is that it is systemic-bias-lite. I'll settle for five years to start most of the articles of interest to those with a fairly parochial view of what constitutes an interesting topic, and 25 years more to catch up with the rest of the planet. You're not telling me that we'll have articles correspording to all the other language versions - total interwiki converage - by 2014?
Charles
2009/2/16 Charles Matthews charles.r.matthews@ntlworld.com:
Thomas Dalton wrote:
2009/2/16 Charles Matthews charles.r.matthews@ntlworld.com:
I believe we have another decade before Wikipedia lives up to its potential as a comprehensive reference. My main hope is that life around the wiki stays dull enough so that the job largely gets done.
Indeed. Current predictions show growth in terms of article numbers pretty much ending in around 4 or 5 years time. We'll then need several more years to actually get all the articles up the scratch. A decade may even be optimistic.
Yeah, well, my reaction to the whole "fruit" discussion is that it is systemic-bias-lite. I'll settle for five years to start most of the articles of interest to those with a fairly parochial view of what constitutes an interesting topic, and 25 years more to catch up with the rest of the planet. You're not telling me that we'll have articles correspording to all the other language versions - total interwiki converage - by 2014?
I'm just going by the statistics, I'm not making any judgements based on anything else. At the moment, we seem to be following a logistic curve which levels out at around 3.5 million articles in around 2013-14. (It's asymptotic, but it will be pretty much there by then.)
2009/2/16 Charles Matthews charles.r.matthews@ntlworld.com:
Yeah, well, my reaction to the whole "fruit" discussion is that it is systemic-bias-lite.
Maybe but that doesn't address the problem. Wikipedia has already reached the point where most people find it includes most of the stuff they carry around in their heads. As a result the average person is facing far fewer opportunities to write new articles or expand existing ones than they used to. This makes both continuing expansion in size and editor numbers somewhat tricky.
On Mon, Feb 16, 2009 at 3:40 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
I'm just going by the statistics, I'm not making any judgements based on anything else. At the moment, we seem to be following a logistic curve which levels out at around 3.5 million articles in around 2013-14. (It's asymptotic, but it will be pretty much there by then.)
So far, "low-hanging fruit" has dominated the growth pattern of Wikipedia. Rather than approaching a horizontal asymptote, we're probably approaching a stable growth rate (i.e., an oblique asymptote), since it's obvious that the number of potential articles yet to be written is not the limiting factor. Rather we're limited by a product of potential articles and users interested in those articles.
But statistically it's probably impossible to know that just from the data, since low-hanging fruit swamps longer-term trends.
-Sage (User:Ragesoss)
This isn't actually accurate. Wikipedia may have reached the point where most people find it includes most of the stuff *that has been traditionally found in encylopedias* they carry around in their heads.
Wikipedia is not paper.
On Mon, Feb 16, 2009 at 3:50 PM, geni geniice@gmail.com wrote:
2009/2/16 Charles Matthews charles.r.matthews@ntlworld.com:
Yeah, well, my reaction to the whole "fruit" discussion is that it is systemic-bias-lite.
Maybe but that doesn't address the problem. Wikipedia has already reached the point where most people find it includes most of the stuff they carry around in their heads. As a result the average person is facing far fewer opportunities to write new articles or expand existing ones than they used to. This makes both continuing expansion in size and editor numbers somewhat tricky.
-- geni
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
2009/2/16 Sage Ross ragesoss+wikipedia@gmail.com:
On Mon, Feb 16, 2009 at 3:40 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
I'm just going by the statistics, I'm not making any judgements based on anything else. At the moment, we seem to be following a logistic curve which levels out at around 3.5 million articles in around 2013-14. (It's asymptotic, but it will be pretty much there by then.)
So far, "low-hanging fruit" has dominated the growth pattern of Wikipedia. Rather than approaching a horizontal asymptote, we're probably approaching a stable growth rate (i.e., an oblique asymptote), since it's obvious that the number of potential articles yet to be written is not the limiting factor. Rather we're limited by a product of potential articles and users interested in those articles.
But statistically it's probably impossible to know that just from the data, since low-hanging fruit swamps longer-term trends.
I think we passed to point where low-hanging fruit was a major factor some time ago (probably round about when we started to level out, although it obviously depends on your definitions). I think in a few years the vast majority of existing topics that we want to include will have at least stubs about them. There will be new topics being created all the time, so growth will never stop completely (there will always be a new series of Big Brother to write about!). We might expand our ideas of what kind of articles are acceptable (ie. relax our notability guidelines), but that's the only way we are going to maintain any significant level of article creation about pre-existing topics.
Thomas Dalton wrote:
2009/2/16 Sage Ross ragesoss+wikipedia@gmail.com:
On Mon, Feb 16, 2009 at 3:40 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
I'm just going by the statistics, I'm not making any judgements based on anything else. At the moment, we seem to be following a logistic curve which levels out at around 3.5 million articles in around 2013-14. (It's asymptotic, but it will be pretty much there by then.)
So far, "low-hanging fruit" has dominated the growth pattern of Wikipedia. Rather than approaching a horizontal asymptote, we're probably approaching a stable growth rate (i.e., an oblique asymptote), since it's obvious that the number of potential articles yet to be written is not the limiting factor. Rather we're limited by a product of potential articles and users interested in those articles.
But statistically it's probably impossible to know that just from the data, since low-hanging fruit swamps longer-term trends.
I think we passed to point where low-hanging fruit was a major factor some time ago (probably round about when we started to level out, although it obviously depends on your definitions). I think in a few years the vast majority of existing topics that we want to include will have at least stubs about them. There will be new topics being created all the time, so growth will never stop completely (there will always be a new series of Big Brother to write about!). We might expand our ideas of what kind of articles are acceptable (ie. relax our notability guidelines), but that's the only way we are going to maintain any significant level of article creation about pre-existing topics.
I think that this was bound to happen; any venture based on describing the known universe has an inherent limit in any case, and it seems obvious that once you've reached some level of coverage, what happens then is more determined by the pace of real life events. However, like software, it's arguable that an encyclopedia is never really finished. Good Articles may be good, and Featured Articles better, but something will always come along to require additions. As for relaxing notability guidelines, I think we very largely get it about right at present, and opening a can of worms does not commend itself to me as a policy.
2009/2/16 Phil Nash pn007a2145@blueyonder.co.uk:
I think that this was bound to happen; any venture based on describing the known universe has an inherent limit in any case, and it seems obvious that once you've reached some level of coverage, what happens then is more determined by the pace of real life events. However, like software, it's arguable that an encyclopedia is never really finished. Good Articles may be good, and Featured Articles better, but something will always come along to require additions. As for relaxing notability guidelines, I think we very largely get it about right at present, and opening a can of worms does not commend itself to me as a policy.
If we get to the point where virtually no new articles are being created (beyond current events) and a very large proportion of the existing articles are at least Good, then it might be worth relaxing the guildlines - what would be the downside? I think a lot of people that like writing new articles don't like the fine tuning that is required to get from Good to Featured, so if we don't let them write new stuff we'll just lose them. We might as well have them doing something.
Another example is that the vast majority of our articles on US Counties have next to nothing about the county history. That is, when was the county formed? What land was it formed out of? Did the boundaries change over time? What was the first city laid out? Who were the first few documentation-attested inhabitants?
Most of our county articles are just current demographics and geographical information. So not only is there a lot of room for new articles, but there is a lot of room for expanding current articles.
In addition a lot of biographical articles on say medieval people are just skeletons showing how they fit into a certain family, with next to nothing about their own life and accomplishments.
Will
2009/2/16 wjhonson@aol.com:
Another example is that the vast majority of our articles on US Counties have next to nothing about the county history. That is, when was the county formed? What land was it formed out of? Did the boundaries change over time? What was the first city laid out? Who were the first few documentation-attested inhabitants?
Most of our county articles are just current demographics and geographical information. So not only is there a lot of room for new articles, but there is a lot of room for expanding current articles.
In addition a lot of biographical articles on say medieval people are just skeletons showing how they fit into a certain family, with next to nothing about their own life and accomplishments.
Oh, absolutely. 5 years is an estimate for when we'll be done creating new articles, it will be a long time after that before we're done enlarging existing articles.
Thomas Dalton wrote:
2009/2/16 Phil Nash pn007a2145@blueyonder.co.uk:
I think that this was bound to happen; any venture based on describing the known universe has an inherent limit in any case, and it seems obvious that once you've reached some level of coverage, what happens then is more determined by the pace of real life events. However, like software, it's arguable that an encyclopedia is never really finished. Good Articles may be good, and Featured Articles better, but something will always come along to require additions. As for relaxing notability guidelines, I think we very largely get it about right at present, and opening a can of worms does not commend itself to me as a policy.
If we get to the point where virtually no new articles are being created (beyond current events) and a very large proportion of the existing articles are at least Good, then it might be worth relaxing the guildlines - what would be the downside? I think a lot of people that like writing new articles don't like the fine tuning that is required to get from Good to Featured, so if we don't let them write new stuff we'll just lose them. We might as well have them doing something.
I think the downside might be exactly what is covered by [[WP:NOT]] at present, and especially [[WP:NOR]]; I've seen several articles that were extremely worthy as research projects, but offended against those policies, and [[WP:SYNTH]] in particular. I hated to nominate for deletion, but it had to be done, within existing policy. I hoped we would not lose those obviously committed and competent editors. The problem is that relaxing [[WP:NOR]] and [[WP:SYNTH]], if not done with extreme care, opens the floodgates to all sorts of abuse, and that is why I don't think it should happen.
As regards quality of articles that pass the intial [[WP:CSD]] and [[WP:N]] tests, it's very largely up to editors being interested enough to dedicate time and effort to take an article to its appropriate level. Take [[Tiddleywink]] as an example; it is notable, by definition, because it's a settlement. But without major research effort, it's extremely unlikely to ever achieve GA, let alone FA. Perhaps that is a problem with the requirements of GA & FA, and perhaps also those criteria are worth looking at.
2009/2/16 Phil Nash pn007a2145@blueyonder.co.uk:
I think the downside might be exactly what is covered by [[WP:NOT]] at present, and especially [[WP:NOR]]; I've seen several articles that were extremely worthy as research projects, but offended against those policies, and [[WP:SYNTH]] in particular. I hated to nominate for deletion, but it had to be done, within existing policy. I hoped we would not lose those obviously committed and competent editors. The problem is that relaxing [[WP:NOR]] and [[WP:SYNTH]], if not done with extreme care, opens the floodgates to all sorts of abuse, and that is why I don't think it should happen.
I think we can relax the notability guidelines without relaxing our fundamental policies. We could change "multiple independent non-trivial sources" to "one independent non-trivial source" (I'm not sure that would allow for many additional articles, but you get the idea).
I wonder about how much of the fruit we've gathered. The plant WikiProject has about 30,000 articles, which include a mixture of articles about plant species, plant morphology and anatomy, and plant biologists. There are close to 300,000 plant species in the world. If we're only in the 5-10% range when it comes to coverage, I could imagine that could easily triple the number of articles without delving into the really hard to find corners. I can imagine that the Arthropods WikiProject (and its daughter projects) have about 13,000 articles under their care. Again, adding 100,000 arthropod articles shouldn't be difficult. True, the stuff that you could add off the top of your head may be gone, but grab a good field guide to plants, or grab a historical dictionary, and you could add hundreds of articles. To me it always seems like time is the major constraint, not stuff that needs to be written about...
On Mon, Feb 16, 2009 at 3:17 PM, Thomas Dalton thomas.dalton@gmail.comwrote:
I think we passed to point where low-hanging fruit was a major factor some time ago (probably round about when we started to level out, although it obviously depends on your definitions). I think in a few years the vast majority of existing topics that we want to include will have at least stubs about them. There will be new topics being created all the time, so growth will never stop completely (there will always be a new series of Big Brother to write about!). We might expand our ideas of what kind of articles are acceptable (ie. relax our notability guidelines), but that's the only way we are going to maintain any significant level of article creation about pre-existing topics.
I'm just starting adding a list of the members of the (US) National Academy of Engineering. we have only about 1% of them covered by articles. there are dozens of fields like that where we haven't even begun on the obvious. We have probably a similar coverage for pre 1990 US state legislators in almost all states.
Additionally, for creative work, for politics, for important products, there will always be subjects for new articles. People are not going to stop doing new things.
What we need is people. We will get them when we stop discouraging new ones by our general hostile way of talking, and our focus on deletion rather than improvement for new articles. Anyone who comes to write even a facebook-type article on themselves is a potential recruit, if talked to carefully. If they're willing to go to the trouble of writing, and brave enough to do so, they can be guided to find a proper subject.
David Goodman, Ph.D, M.L.S. http://en.wikipedia.org/wiki/User_talk:DGG
On Mon, Feb 16, 2009 at 7:56 PM, George Herbert george.herbert@gmail.com wrote:
On Mon, Feb 16, 2009 at 11:42 AM, Thomas Dalton thomas.dalton@gmail.comwrote:
We're shrinking because we've already written most of the stuff we want to include.
This is orthogonal to the main conversation here, but this is not nearly the case.
We've picked off a lot of low hanging fruit, approaching all of it. Things which haven't been dealt with include [...]
<snip>
I wondered why this thread had exploded with activity. It's because it turned into a "low hanging fruit" debate!
My approach to seeing how comprehensive Wikipedia's coverage is at the moment is, while reading a book or watching a TV documentary, to mentally make notes of things to look up on Wikipedia. I did that yesterday while watching "The Victorians" (a BBC documentary presented by Jeremy Paxman where he looked at the Victorians through their paintings).
There was lots I could have looked up, including the program itself (no article, understandably enough, as it wouldn't have met notability guidelines), but the three things I made a mental note of were:
Gustave Dore:
http://en.wikipedia.org/wiki/Gustave_Dore
Manchester Town Hall:
http://en.wikipedia.org/wiki/Manchester_Town_Hall
1888 International Exhibition in Glasgow:
http://en.wikipedia.org/wiki/Glasgow_Festivals#Past_Festivals
The first two had articles, but the third one doesn't have its own article. Turns out there are three big exhibitions that were held in Glasgow, in 1888, 1901 and 1911 that we don't have articles on.
We do have one on the one in 1938:
http://en.wikipedia.org/wiki/Empire_Exhibition,_Scotland_1938
And the Garden Festival in 1988:
http://en.wikipedia.org/wiki/Glasgow_Garden_Festival
But it's the historical stuff that hasn't been written about yet (and that's not even mentioning the art history - I should have noted the titles of all the artworks and the artist's and seen which we had articles on).
I was kind of hoping that an interesting set of murals in the Manchester Town Hall hadn't had an article written on them yet, but it has been fairly well covered already:
http://en.wikipedia.org/wiki/The_Manchester_Murals
History is an almost boundless area for new articles.
Another way to assess how comprehensive Wikipedia is, is to take some document (or even one of our unwikified articles) and wikify it in some reasonably sensible way and see how many of the links are red. This is a bit more exciting than wikifying some index or list of entries in an old encyclopedia (though the latter is a more efficient way to do this sort of thing).
One other thing that people sometimes forget to do is to check "what links here" for said redlinks and see how popular they are. See how many other people have been trying to link to it. Though you have to remember to do a search as well and pick up the plain text examples of the redlinked article that haven't been linked (some of which should be, some shouldn't).
It's very satisfying to write a new article that has 10 or so incoming links already! :-)
Carcharoth
2009/2/16 Guettarda guettarda@gmail.com:
True, the stuff that you could add off the top of your head may be gone, but grab a good field guide to plants, or grab a historical dictionary, and you could add hundreds of articles. To me it always seems like time is the major constraint, not stuff that needs to be written about...
Those sources will give you stubs, will they give you much more? I guess it depends on how specific a field guide you have. (A general "Field Guide to Plants" won't give you more than a couple of lines, a "Field Guide to Orchids" might give you a paragraph or two, a "Field Guide to Orchids of the British Isles" might give you most of an article.)
On Tue, Feb 17, 2009 at 3:31 AM, Thomas Dalton thomas.dalton@gmail.com wrote:
Those sources will give you stubs, will they give you much more? I guess it depends on how specific a field guide you have.
Stubs aren't bad things.
Plus, the general guide can give you things to then look up in more depth elsewhere.
-Matt
2009/2/17 Matthew Brown morven@gmail.com:
On Tue, Feb 17, 2009 at 3:31 AM, Thomas Dalton thomas.dalton@gmail.com wrote:
Those sources will give you stubs, will they give you much more? I guess it depends on how specific a field guide you have.
Stubs aren't bad things.
Indeed, but there are far more topics that it is easy to write a stub about than there are topics that it is easy to write a whole article about.
Plus, the general guide can give you things to then look up in more depth elsewhere.
Yes, but once you're using one source to find other sources and hunting for them, you're not really in the realms of low-hanging fruit.
On Tue, Feb 17, 2009 at 1:52 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
<snip>
Yes, but once you're using one source to find other sources and hunting for them, you're not really in the realms of low-hanging fruit.
Some of the so-called low-hanging fruit are articles that have never been in that good condition, even now, or that still have great potential for expansion or reorganisation (even if a lot of the detail is in related articles, accessible via links). I'll try and find a few examples.
http://en.wikipedia.org/wiki/Sky
Compare that to:
http://en.wikipedia.org/wiki/Soil
...which is quite good.
The surgery article is interesting:
http://en.wikipedia.org/wiki/Surgery
...but quite why there are 15 templates at the bottom of the article, I don't know.
Something that looks OK at first glance, but less so when you look closer, is:
http://en.wikipedia.org/wiki/Grass
Another article that is in a "hodge-podge" state is:
http://en.wikipedia.org/wiki/Begging
If these all count as low-hanging fruit, they may have been picked, but they haven't really ripened yet. Part of the trouble is that truly general, overview articles are: (a) difficult to write well; and (b) experts tend to prefer to write more limited, specialised articles. Sometimes the subsidiary articles need to be written to a good level before the general article can be tackled. Sometimes it is the other way round.
Carcharoth
2009/2/17 Carcharoth carcharothwp@googlemail.com:
If these all count as low-hanging fruit, they may have been picked, but they haven't really ripened yet. Part of the trouble is that truly general, overview articles are: (a) difficult to write well; and (b) experts tend to prefer to write more limited, specialised articles. Sometimes the subsidiary articles need to be written to a good level before the general article can be tackled. Sometimes it is the other way round.
I love a good metaphor! You're absolutely right, writing articles about very general topics is very difficult. I think the problem comes in trying to balance breadth, depth and conciseness. There isn't an obvious solution - it just needs someone with the right talents to devote a significant amount of time and effort to it.
2009/2/17 Carcharoth carcharothwp@googlemail.com:
If these all count as low-hanging fruit, they may have been picked, but they haven't really ripened yet. Part of the trouble is that truly general, overview articles are: (a) difficult to write well; and (b) experts tend to prefer to write more limited, specialised articles. Sometimes the subsidiary articles need to be written to a good level before the general article can be tackled. Sometimes it is the other way round.
I've noticed that featured articles are rarely general topics - they tend to be specialised articles brought to FA status by one person interested in that specialist subtopic.
- d.
On 17/02/2009, Thomas Dalton thomas.dalton@gmail.com wrote:
I love a good metaphor! You're absolutely right, writing articles about very general topics is very difficult. I think the problem comes in trying to balance breadth, depth and conciseness. There isn't an obvious solution - it just needs someone with the right talents to devote a significant amount of time and effort to it.
I believe that there's probably a class of general articles that will never go FA. Some topics will not fit into the combination of guidelines (particularly length), and to the extent that these guidelines are rigidly enforced, then you have no chance.
The only way to get some articles (particularly general ones) past FA is to remove stuff from them, and hope nobody in the FA review notices ;-)
On Tue, Feb 17, 2009 at 2:39 PM, David Gerard dgerard@gmail.com wrote:
2009/2/17 Carcharoth carcharothwp@googlemail.com:
If these all count as low-hanging fruit, they may have been picked, but they haven't really ripened yet. Part of the trouble is that truly general, overview articles are: (a) difficult to write well; and (b) experts tend to prefer to write more limited, specialised articles. Sometimes the subsidiary articles need to be written to a good level before the general article can be tackled. Sometimes it is the other way round.
I've noticed that featured articles are rarely general topics - they tend to be specialised articles brought to FA status by one person interested in that specialist subtopic.
I've looked briefly through the list at WP:FA (briefly because there are 2,419 of them), and some that strike me as particularly general (and nearly all are common terms) are:
http://en.wikipedia.org/wiki/Seabird http://en.wikipedia.org/wiki/Ant http://en.wikipedia.org/wiki/Bacteria http://en.wikipedia.org/wiki/Bird http://en.wikipedia.org/wiki/Dinosaur http://en.wikipedia.org/wiki/Frog http://en.wikipedia.org/wiki/Primate http://en.wikipedia.org/wiki/Virus http://en.wikipedia.org/wiki/Actuary http://en.wikipedia.org/wiki/Diamond http://en.wikipedia.org/wiki/Welding
Less general, but still very broad, are ones like:
http://en.wikipedia.org/wiki/DNA http://en.wikipedia.org/wiki/Economy_of_India http://en.wikipedia.org/wiki/Electrical_engineering http://en.wikipedia.org/wiki/Medieval_cuisine
I stopped at "Food and drink" as we have quite a lot of featured articles.
One that I had in mind as an example is now a former featured article:
http://en.wikipedia.org/wiki/Invasion
Looking through the lists of former featured articles will probably yield similar examples.
One thing I did discover is that reading through the list at WP:FA is really quite difficult.
It is 29 sections, but each section is a wall of text and it is quite hard to browse. I might try and do a personalised listing at some point, bringing out the areas I'm interested in and slicing up the FA cake in a different way. Such as identifying the more "general" ones and the more "niche" ones, and the "specific" items such as games, films, books, events, and paintings (as opposed to genres, histories and stuff like that), and biographies and suchlike. But with so many articles, it's difficult to do that.
Carcharoth
On Tue, Feb 17, 2009 at 11:05 AM, Carcharoth carcharothwp@googlemail.com wrote:
I might try and do a personalised listing at some point, bringing out the areas I'm interested in and slicing up the FA cake in a different way. Such as identifying the more "general" ones and the more "niche" ones, and the "specific" items such as games, films, books, events, and paintings (as opposed to genres, histories and stuff like that), and biographies and suchlike. But with so many articles, it's difficult to do that.
That would be interesting. I wonder if this could be something that could be integrated into the 1.0 rating scheme... another, parallel rating for "scope" or "generality". Naturally, any such determinations will be subjective, but so are article ratings and yet the semi-codified Stub-Start-C-B ratings tend to work out pretty well. It would be great to have the breakdown of general vs. specific articles not just for FAs, but for everything.
-Sage (User:Ragesoss)
2009/2/17 Sage Ross ragesoss+wikipedia@gmail.com:
That would be interesting. I wonder if this could be something that could be integrated into the 1.0 rating scheme... another, parallel rating for "scope" or "generality". Naturally, any such determinations will be subjective, but so are article ratings and yet the semi-codified Stub-Start-C-B ratings tend to work out pretty well. It would be great to have the breakdown of general vs. specific articles not just for FAs, but for everything.
That might be good. It would also help when determining if an article being an orphan is a problem. Very specific articles probably won't be linked to much, more general articles will be. So, if a general article is an orphan, we have a problem, if a specific article is, we probably don't.
On Tue, Feb 17, 2009 at 5:25 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
2009/2/17 Sage Ross ragesoss+wikipedia@gmail.com:
That would be interesting. I wonder if this could be something that could be integrated into the 1.0 rating scheme... another, parallel rating for "scope" or "generality". Naturally, any such determinations will be subjective, but so are article ratings and yet the semi-codified Stub-Start-C-B ratings tend to work out pretty well. It would be great to have the breakdown of general vs. specific articles not just for FAs, but for everything.
That might be good. It would also help when determining if an article being an orphan is a problem. Very specific articles probably won't be linked to much, more general articles will be. So, if a general article is an orphan, we have a problem, if a specific article is, we probably don't.
Rudimentary suggestions based on searches can probably generate suggestions for links for practically any article. Humans could then go through those lists working out if links are needed. If you are just presented with an orphaned article, it can be a pain trying to work out where it can be linked from.
To take an example, both of a low-hanging fruit and a relatively orphaned article:
http://en.wikipedia.org/wiki/Lister_Medal
I was very surprised, back in October 2008, to discover that we didn't have an article on this prestigious award made to surgeons. So I created a list of those awarded the medal. Turns out the awarding institute don't have a handy list on their website, so that was probably the reason the article hadn't been created, but that's not the point I'm making here. The point I'm making is that I failed to link it from anywhere very much.
About 4 months later, it's still not linked from anywhere much:
http://en.wikipedia.org/w/index.php?title=Special:WhatLinksHere/Lister_Medal...
* Joseph Lister, 1st Baron Lister (links) * Manchester Mark 1 (links) * Regius Professor of Surgery, Glasgow (links)
The first link, from Lister's article, was added by me in October 2008:
http://en.wikipedia.org/w/index.php?title=Joseph_Lister,_1st_Baron_Lister&am...
The other two links were added as follows:
http://en.wikipedia.org/w/index.php?title=Regius_Professor_of_Surgery,_Glasg...
That third link was added with the creation of that article in December 2008 (on a side-note, that list of Regius Professors of Surgery should be redlinks, not bare text, but the problem is that at least four of them are blue links to the wrong articles - this is where redlinks don't always work so well, unless the disambiguation naming is obvious enough).
http://en.wikipedia.org/w/index.php?title=Manchester_Mark_1&diff=2698668...
The second link (diff above) was added in February 2009. It was piped, though in fact a redirect from Lister Oration already existed.
But two links in four months seems pretty poor to me. Or is it?
What I should have done when I created the article, and what I will do at some point (if no-one else does it first), is go through the articles of the medallists linking back to the medal (and adding sources), and do a search for the various terms (Lister Medal, Lister Oration), and link them from various articles. In this case, there isn't much mention of the medal in other articles, but for other orphaned articles there can be.
And I should finish writing the article as well. It's still pretty stubby.
Carcharoth
Charles Matthews wrote:
Jussi-Ville Heiskanen wrote:
In that sentence there are buried assumptions as follows:
- There are people on wikipedia who will not permit
quality.
People who won't permit quality are aggressive.
There is a clear unambiguous metric for quality.
Aggressive people who won't permit quality will
follow an article.
- Over the long term, the dynamics of wikipedias
practices will not prevent editors who will not allow quality on wikipedia from dragging it down to the level that they aggressively and persistently insist on bringing it down to. There are no working heuristics to allow it to transcend that attractor.
*Understanding* the logical flaws of those 5 statements is left to the student.
It would be rash to say you couldn't find any examples where this is true - there is a large selection of articles. It might be a fair model for the article about, for example, a controversial Governor of Alaska who didn't get chosen as a candidate for Vice-President. But you could click Random Article for a little while before you came up with an article to which this argument really would apply.
I disagree that even Mrs. Palins article will fulfill the claim of my 5th paraphrase.
Long term, (think 5 years down the line, or even say twice the current age of wikipedia itself) those little problems will be transcended.
Yours,
Jussi-Ville Heiskanen
Thomas Dalton wrote:
2009/2/17 Matthew Brown morven@gmail.com:
On Tue, Feb 17, 2009 at 3:31 AM, Thomas Dalton thomas.dalton@gmail.com wrote:
Those sources will give you stubs, will they give you much more? I guess it depends on how specific a field guide you have.
Stubs aren't bad things.
Indeed, but there are far more topics that it is easy to write a stub about than there are topics that it is easy to write a whole article about.
Erk... this is what we have the template {{notastub}} for.
"A short article is not a stub." Repeat 10 times under your breath.
<sarcastic aside> Otherwise, why would the 1975 Encyclopaedia Britannica Micropaedia article on "Monastery" consist of 12 words? </sarcastic aside>
But completely seriously, a subject that can be exhaustively covered briefly, is not a stub. Period.
Yours,
Jussi-Ville Heiskanen
2009/2/19 Jussi-Ville Heiskanen cimonavaro@gmail.com:
Thomas Dalton wrote:
2009/2/17 Matthew Brown morven@gmail.com:
On Tue, Feb 17, 2009 at 3:31 AM, Thomas Dalton thomas.dalton@gmail.com wrote:
Those sources will give you stubs, will they give you much more? I guess it depends on how specific a field guide you have.
Stubs aren't bad things.
Indeed, but there are far more topics that it is easy to write a stub about than there are topics that it is easy to write a whole article about.
Erk... this is what we have the template {{notastub}} for.
"A short article is not a stub." Repeat 10 times under your breath.
<sarcastic aside> Otherwise, why would the 1975 Encyclopaedia Britannica Micropaedia article on "Monastery" consist of 12 words? </sarcastic aside>
But completely seriously, a subject that can be exhaustively covered briefly, is not a stub. Period.
All true, but not at all relevant to what I was saying. I was talking about how much you can *easily* write about a subject, not how much there is to write about it.
On Wed, Feb 18, 2009 at 6:14 PM, Jussi-Ville Heiskanen cimonavaro@gmail.com wrote:
But completely seriously, a subject that can be exhaustively covered briefly, is not a stub. Period.
Yes. A stub is a place-holder for a real article. I suspect that the name is by analogy with a stub function in programming, since the people who started Wikipedia included many programmers; a stub in that context is a do-nothing or test-data implementation of a function, created so the project as a whole can be compiled and the parts that aren't the stub can be given some testing.
Thus, a stub on Wikipedia is something that's there INSTEAD of the proper article that should be at that location. People create them so they can categorize them, link them in properly, put a few appropriate 'see also', 'external links' and sources, and a talk page, before they have time or other resources to write a real encyclopedia article there.
There are plenty of things in the world that everything that should be said about them fits in a paragraph. This doesn't make them not worth an article.
-Matt
On Sun, Feb 15, 2009 at 11:27 PM, Carl Beckhorn cbeckhorn@fastmail.fm wrote:
Regardless of the history, Sanger does have a viewpoint that would be worth reading even if the author were anonymous.
Only, he does not feel this way about the viewpoints of others who are anonymous.
On Mon, Feb 16, 2009 at 9:53 AM, Carcharoth carcharothwp@googlemail.com wrote:
So maybe I should have read that draft instead. It would be nice to know which versions were approved by the three editors above, and at what stage.
Ah, be patient, Carch. Since enwiki rejects FlaggedRevs as antithetical to open editing, I predict Larry will pick up on it as it affords Sangerpedia a cheap, trivial way to be Radically Different from Jimbopedia.
Actually using the tool, to tighten up the status quo which he considers Still Too Open (to dissent, for example), will just be a pleasant side effect. The same can be said about knowing who approved which edits, this helps those studying the editorial forensics of a failing project but it is still secondary to creating a deep philosophical contrast.
On Mon, Feb 16, 2009 at 2:40 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
I'm just going by the statistics, I'm not making any judgements based on anything else. At the moment, we seem to be following a logistic curve which levels out at around 3.5 million articles in around 2013-14.
"The end is near!" "Which end?"
In breadth of coverage Wikipedia is still in its early adolescence. Myself I learned a lesson about guessing numbers—don't bother, sweet chariot, you'll always swing too low.
On Mon, Feb 16, 2009 at 3:52 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
I think a lot of people that like writing new articles don't like the fine tuning that is required to get from Good to Featured
I don't know about all that. When I write a new article I don't like the pedantic ref-bombing that is needed to prevent it from being deleted 16.9 seconds later... but I still do it... to hell with the other stuff.
—C.W.
On Mon, Feb 16, 2009 at 3:40 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
I'm just going by the statistics, I'm not making any judgements based on anything else. At the moment, we seem to be following a logistic curve which levels out at around 3.5 million articles in around 2013-14. (It's asymptotic, but it will be pretty much there by then.)
Is there any data on changes in the percentage of work spent on adding new material vs. undoing damage to existing material? I'm thinking not only of vandalism, but "clueless edits": people posting religious evangelism, pushing pet theories, adding bogus facts to make their country/city/whatever look more significant than it really is, replacing good writing with bad writing, etc.
Hypothesis: The more good material there is, the more human effort it takes to keep it from getting degraded. So, nearing the asymptote, most serious Wikipedia editors may end up spending most of their time doing reverts. An unpleasant thought.
Ben
2009/2/21 Ben Kovitz bkovitz@acm.org:
On Mon, Feb 16, 2009 at 3:40 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
I'm just going by the statistics, I'm not making any judgements based on anything else. At the moment, we seem to be following a logistic curve which levels out at around 3.5 million articles in around 2013-14. (It's asymptotic, but it will be pretty much there by then.)
Is there any data on changes in the percentage of work spent on adding new material vs. undoing damage to existing material? I'm thinking not only of vandalism, but "clueless edits": people posting religious evangelism, pushing pet theories, adding bogus facts to make their country/city/whatever look more significant than it really is, replacing good writing with bad writing, etc.
I've seen pie charts showing what proportion of edits are reverted, reverts and genuine article improvements. I can't remember where, though...
Hypothesis: The more good material there is, the more human effort it takes to keep it from getting degraded. So, nearing the asymptote, most serious Wikipedia editors may end up spending most of their time doing reverts. An unpleasant thought.
That's a very interesting point... FlaggedRevs may help there - if it's turned on for the entire site it would allow for more efficient RC patrol. Even with that, we may eventually hit a point where there is too much vandalism to cope with (at least with FlaggedRevs we would have a very clear metric - the age of the oldest unreviewed edit), at which point we may have to take unpleasant measures (banning anonymous editing is the obvious one - I hope we never get to the point where that is necessary...).
On Feb 16, 2009, at 12:20 AM, Tim Starling wrote:
"Sanger was one of the founders of Wikipedia, and of its failed predecessor Nupedia, who left the fold because of differences over the question of the proper role of experts."
Strange, I thought it was because he stopped being paid for it.
That's right. (As your next quotation showed.)
He then forgot about Wikipedia completely for a few years, and re-emerged a critic once the media started paying attention to it.
I think this does Larry Sanger an injustice, at least insofar as it suggests that he's some sort of media whore. From the beginning, he wanted credentialed experts to compile the "real" encyclopedia, using the general-public wiki as a "feeder". He has strenuously made this exact same objection--that letting just anyone edit the live encyclopedia would introduce bias in favor of "idiots" with hobby horses--from the very beginning, even before the project officially started.
Ben
On Feb 16, 2009, at 2:10 PM, Thomas Dalton wrote:
We could discuss why [CZ] failed but I think the real answer is simply that Wikipedia is "good enough" so there is very little interest in a new project doing the same thing.
I think you have pegged it exactly right. In most large markets, the rule of thumb is that the #1 player holds 40% market share, the #2 player holds 20% market share, the #3 player holds 10%, and then there are some little guys. Most markets on the Internet, though, are "winner take all". For example: eBay, Amazon, Wikipedia. It's very hard for a newcomer to displace an established top player, *regardless of quality* (unless the quality difference is revolutionary). So, we will likely never be able to test Larry Sanger's claim that giving experts ultimate say would produce a better encyclopedia.
Disclaimer: I've read three or four books on marketing, so that, uh, makes me an expert. ;)
Ben
2009/2/21 Ben Kovitz bkovitz@acm.org:
Disclaimer: I've read three or four books on marketing, so that, uh, makes me an expert. ;)
Probably more of an expert than someone who's read ten or twenty books on marketing ;)
2009/2/21 James Farrar james.farrar@gmail.com:
2009/2/21 Ben Kovitz bkovitz@acm.org:
Disclaimer: I've read three or four books on marketing, so that, uh, makes me an expert. ;)
Probably more of an expert than someone who's read ten or twenty books on marketing ;)
I've read the Wikipedia article on marketing, what more could I need?
2009/2/21 Thomas Dalton thomas.dalton@gmail.com:
2009/2/21 James Farrar james.farrar@gmail.com:
2009/2/21 Ben Kovitz bkovitz@acm.org:
Disclaimer: I've read three or four books on marketing, so that, uh, makes me an expert. ;)
Probably more of an expert than someone who's read ten or twenty books on marketing ;)
I've read the Wikipedia article on marketing, what more could I need?
http://en.citizendium.org/wiki/Marketing , obviously.
- d.
2009/2/21 David Gerard dgerard@gmail.com:
2009/2/21 Thomas Dalton thomas.dalton@gmail.com:
2009/2/21 James Farrar james.farrar@gmail.com:
2009/2/21 Ben Kovitz bkovitz@acm.org:
Disclaimer: I've read three or four books on marketing, so that, uh, makes me an expert. ;)
Probably more of an expert than someone who's read ten or twenty books on marketing ;)
I've read the Wikipedia article on marketing, what more could I need?
http://en.citizendium.org/wiki/Marketing , obviously.
I would rather take my marketing advice from an encyclopaedia people read...
http://www.xkcd.com/125/ (see the tooltip text)
On Sat, Feb 21, 2009 at 9:24 AM, Ben Kovitz bkovitz@acm.org wrote:
On Feb 16, 2009, at 2:10 PM, Thomas Dalton wrote:
We could discuss why [CZ] failed but I think the real answer is simply that Wikipedia is "good enough" so there is very little interest in a new project doing the same thing.
I think you have pegged it exactly right. In most large markets, the rule of thumb is that the #1 player holds 40% market share, the #2 player holds 20% market share, the #3 player holds 10%, and then there are some little guys. Most markets on the Internet, though, are "winner take all". For example: eBay, Amazon, Wikipedia. It's very hard for a newcomer to displace an established top player, *regardless of quality* (unless the quality difference is revolutionary). So, we will likely never be able to test Larry Sanger's claim that giving experts ultimate say would produce a better encyclopedia.
Is Amazon really in a "winner takes all" market? My intuition, as well an incredibly quick search, suggests "no way" ( http://toc.oreilly.com/2008/04/amazon-growth-fuels-onlines-bo.html says "while Amazon is by far the most dominant player, they still represent less than half of the total online book market"). If you've got stats that say otherwise though, I'd gladly accept them over some random guy on the Internet.
eBay I can understand, but I see no reason why Wikipedia will be a winner takes all market. All that's needed is a decent search engine for educational content, which doesn't focus so much on popularity. I can't imagine that not happening one of these days.
Maybe someone should write a typical history of which policies new users tread on. http://en.wikipedia.org/wiki/User_talk:Ken_Birman That guy is an expert, because he actually wrote software, and he is a professor at Cornell. [[gossip protocol]] and [[virtual synchrony]] are his pet projects. I found gossip protocol to be something like Artificial Neural Networks in 1990. Everyone was talking about them. Everyone had different ideas on what it means. (Still do). (Artificial Logic Networks) ALN's are a close cousin to ANN's. If you look on the history of his talk page, then you will see that it has shaped up more like a blog, which is at one point I could not find a welcome notice of any kind on his talk page, then there was some critique from KPBotany. Since I could not find a welcome notice, I put one up. That stuff ain't there, so maybe he was following my lead about how to maintain a talk page. I also told him about my {{prod}} for deleting [[gossip protocol]], saying at one point that he had a defender.
I was not happy with what little that defender would learn from [[gossip protocol]], because it seemed to be ultimate -- optimal in one case and understandably short of optimal in practice. So, I went to work, trying to understand the math behind potentials, and lo and behold I found I had already written! evidence of "optimal" (exponential propagation) and "worst case" (everyone talking at once, which yields "in proportion to number of participants"). I needed a monographic example. I think Ken has decided to stay out of [[gossip protocol]], as long as it does not get deleted. I do not think he will decide to write my monographs into what is mostly his work.
Tthe role of experts IN wikipedia is to learn from experts ON wikipedia. {{subst:uw-notcensored1}} Coulda fooled me, and I do not feel like arguing on either side.