Paul Campos has the story (following up on here, following up on here), and it makes wikipedia look pretty bad. It’s a rabbit hole within a rabbit hole within a rabbit hole. Rabbit holes all the way down.
On the other hand, Wikipedia is wonderful; indeed, it was through wikipedia that we were able to find some of the information used to find holes in the Fred Bonine story.
And, speaking as a matter of governance, it makes sense that (a) it should be easy for people to add material to wikipedia pages, and (b) it should be hard for people to take material out.
Combining these two reasonable features—ease of entry and stickiness—inevitably will lead to errors that are hard to remove.
That said, I expect that in short order the Fred Bonine entry will get cleaned up. And wikipedia is about a zillion times better than the scientific literature (or, for that matter, news media and social media) in correcting mistakes when they are clearly labeled. The NYT never corrected those David Brooks and Nicholas Kristof columns, Freakonomics never corrected its endorsements of junk science, Ted, PNAS, NPR, etc etc etc. Wikipedia has errors too, and there’s a problem when people take Wikipedia as Truth, but I still think its error correction is better than just about any other institution out there.
Wikipedia does have a preference for what it calls reliable sources, and also for secondary sources over primary sources. This means that a NY Times article about the raw data is preferred over the raw data itself, and over original research about that raw data. This rule has its merits, but it does sometimes mean that obvious errors do not get corrected.
Wikipedia prefers *secondary* sources? As practice or as policy? That just seems backwards. And doesn’t line up with my experience with wikipedia, but I’ve only dipped my toe in the water, so it’s possible I just haven’t encountered it.
Yes, Wikipedia frowns upon primary sources. They don’t want Wikipedia editors to, say, read a bunch of participants’ diaries and letters, and from that construct an account of a historical event– that would be original research, and one of Wikipedia’s rules is “No Original Research”. An analysis by a historian would be preferable to contemporary newspapers, but a newspaper account is preferable to what a Wikipedia editor can find by rummaging around in the archives.
This is not always straightforward– is Herodotus a primary or secondary source?– but that’s what the preferences are.
Wikipedia explicitly prefers secondary sources, and I and others who have tried to link to the only source for a claim have been reverted and told to find some published source summarizing that source. They even have a standard warning: “This page cites too many primary sources! Improve it by adding more references to secondary sources.” Likewise, if I publish a blog post under my professional name which cites primary sources or shows the problems in a published source, I can’t cite that- but if I send it to some kind of magazine or journal, it magically becomes citable on Wikipedia.
It is worth studying Wikipedia’s rules on what can be cited, because they are utterly bizarre if you come from an academic background or a journalism background. And they create a situation where people with too much time on their hands can keep control of an article just by memorizing Wikipedia’s bizarre rules and raising procedural objections to changes they don’t like. German Wikipedia does not seem bad, and some kinds of topics are not bad, but for anything which requires judgement or is tied to large political movements- oh howdy is Wikipedia bad.
+1
Sure the rules are somewhat arcane, but pretty sensible if you consider that Wikipedia by design does not try to evaluate the quality of primary research. It is explicitly not an authority or purport to have hired or vetted relevant experts to monitor submissions. And that is what would be needed if they were to start accepting original research. Sure, you may actually be an expert in your area of research, but there are plenty of cranks (even cranks with PhDs!) who aren’t taken seriously by their peers, and as such, are even *more* motivated to contribute their nonsense to Wikipedia.
Respected journals, magazines, newspapers, etc. pay professionals to maintain basic standards of quality of research. And think about how often they get it wrong. I can’t imagine that Wikipedia would be better off allowing the highly motivated cranks to submit their own research and attempt to have unpaid, amateur Wikipedia editors adjudicate.
The thing is, who exactly is a respected source? If you have an academic paper on a subject, which is then misreported on by a mainstream news source, who does Wikipedia believe?
Frankly, I don’t think wikipedia editorial policy does very much to deter cranks. It’s quite easy to get crankery past the rules. It’s really more about maintaining what a certain small cabal considers the essential elements of wikipedia’s respectability.
I have noticed they typically have no idea what they are talking about. This has been going on for a long time given Gell-man amnesia effect was coined in 2002.
https://en.m.wikipedia.org/wiki/Michael_Crichton#GellMannAmnesiaEffect
Then there is this classic comic: https://phdcomics.com/comics/archive/phd051809s.gif
My primary use of wikipedia is as a source of interesting links about a topic. So the no primary sources thing really reduces the usefulness, with dubious benefit.
All I can say is that the Wikipedia rules contradict basic principles of both journalistic epistemology and academic epistemology, and that because they are so idiosyncratic, people with too much time on their hands can dominate an article by internalizing the arcane rules.
If there is only one sources for a claim such as Bill Thayer’s example, I can’t see anything better than quoting that source. The article could then go on to summarize what people have said about that source, but it would need to say when necessary “but this contradicts the original source.”
“…if I publish a blog post under my professional name which cites primary sources or shows the problems in a published source, I can’t cite that”
Yes, you can. Maybe. Your blog post that cites primary sources is a secondary source, so you’ve already passed that bar. However, it’s not *sufficient* that your blog post is secondary; there is an additional requirement that it be a *reliable* secondary source. Any blog post that is self-published, which is to say, you can upload whatever you want without peer-review or editorial oversight is considered unreliable ipso facto, and therefore you could not cite your blog post, even though it may be a secondary source. If you upload a story about watching a guy win the world bungee jumping contest, you cannot cite your blog, because it’s primary. If the guy standing next to you works for the Forgottenville Gazette, writes up the almost identical story that you just posted on your blog, and publishes it in the paper, you can then write about it on Wikipedia, citing the article the guy wrote.
Bill Thayer has one example of how editors can use Wikipedia policies to keep nonsense in and experts out https://rambambashi.wordpress.com/2012/09/14/original-research-at-wikipedia/ Pretty much anyone I have talked to who has tried to substantively edit Wikipedia on a topic they are an expert on has similar stories.
I once tried to get a Wikipedia page on math fixed so that it didn’t contain a lot of wrong statements. I didn’t want to just change the page because I figured the people who wrote the stuff would just change it back. So, I explained to them why what they wrote was wrong. They insisted it was right (one was more insistent than the other). I eventually gave up. Of course, anyone can edit Wikipedia, but that works both ways.
About 10 years ago I rewrote nearly an entire page on a biomed topic. I even created some nice figures. Even though now I am much more skeptical about all that, it still conveys the general belief/understanding.
It is somewhat satisfying to see most of it unchanged. I see the figures have been translated into like 10 other languages and one even included in some kind of online textbook.
So sometimes it works out, other times it does not. I’m surprised that a math page is controversial though.
Technical math is generally fine. I did fix the page on L’Hôpital’s rule once, but I see they’ve messed it up again: the hypotheses are not the way a mathematician would state them, i.e., they are redundant (but less liable to be misinterpreted). However, for the things that all mathematicians know, Wikipedia is not reliable.
In the page for L’Hopital’s rule, the requirement that g’ 0 near c is implied by the limit of f/g existing. The paragraph that starts “The hypothesis that g'(x) 0 appears most commonly in the literature …” is nonsense. Just because wrong things appear in some “literature” doesn’t mean you should repeat them.
The “Equality (mathematics)” article is wrong, but perhaps slightly less wrong than in the past.
The “Mathematical proof” article appears (on skimming) to be slightly better than in the past, but still seems to use so many words that it will obscure more than enlighten. I doubt someone who doesn’t know what a proof is will be able to tell you what it is after reading that page.
In the page for L’Hôpital’s rule, the requirement that g’ 0 near c is implied by the limit of f/g existing. The paragraph that starts “The hypothesis that g'(x) 0 appears most commonly in the literature …” is nonsense. Just because wrong things appear in some “literature” doesn’t mean you should repeat them.
The “Equality (mathematics)” article is wrong, but perhaps slightly less wrong than in the past.
The “Mathematical proof” article appears (on skimming) to be slightly better than in the past, but still seems to use so many words that it will obscure more than enlighten. I doubt someone who doesn’t know what a proof is will be able to tell you what it is after reading that page.
I’m a fan of the co-founder Larry Sanger’s new project which he is calling the Encyclosphere. The goal is a more decentralized version of Wikipedia where there are multiple versions of the same page maintained by different groups. The prominence of Wikipedia has lead to an entire industry forming around it (along with the differing incentives of hobbyists vs actual experts editing it). An example is this PR firm which focuses on wikipedia: https://wiki-pr.com/
There are people in the above thread saying that wikipedia doesn’t allow primary sources. I guess it depends on the context. I was curious so I went to wikipedia article on Mister P, and it had lots of primary sources, including our original 1997 article.
I think what it really is is that Wikipedia doesn’t want to BE the primary source. If you do a bunch of research and claim that Bonine was promoted fraudulently as a world record holder by boxing promoters… You can make that claim on a blog post or a journal article or a detailed letter to your local newspaper or whatever, but you can’t “publish an article” on wikipedia itself in which you lay out the evidence for the claim.
What wikipedia doesn’t like is “original research.” I believe this includes compilation and interpretation of primary sources without some kind of credentialed authority.
https://en.wikipedia.org/wiki/Wikipedia:No_original_research#Primary,_secondary_and_tertiary_sources
Yes Daniel, That makes sense
“I went to wikipedia article on Mister P, and it had lots of primary sources, including our original 1997 article.”
The Mister P page is given a quality rating of “start,” implying there is lots more work to do. I suspect it got this rating for only one reason, and that is too much reliance on primary sources.
It is a good poster child for a topic also addressed in this thread, which is that changing it so that it relies on secondary sources will definitely dumb it down.
That being said, I am also a supporter of wikipedia and despite its warts, I can no longer envision any improvements from changes in policy, they are adequate the way they are.
It also depends on whether a stickler for the rules has noticed that page. German Wikipedia seems much more scholarly than English Wikipedia, and I suspect they have a culture of not placing too much emphasis on the “no primary sources” policy. And on any Wikipedia, a lot depends on what experts in their policies want a page to look like. If there is only one person who knows “Wikipedia speak” and cares about a page, that one person will win most edit wars because he or she can cite Wikipedia policies better than the average editor.
Contrary to many of the claims made in the comments here, there is no Wikipedia policy stating that you can’t use primary sources. The policy in question only says that secondary sources should be preferred in most cases, because they are less prone to cherry-picking of sources or misinterpretation by editors:
“Wikipedia articles should be based on reliable, published secondary sources and, to a lesser extent, on tertiary sources and primary sources. Secondary or tertiary sources are needed to establish the topic’s notability and avoid novel interpretations of primary sources”
(“https://en.wikipedia.org/wiki/Wikipedia:No_original_research#Primary,_secondary_and_tertiary_sources”).
And nevertheless such a policy is enforced eg. this screenshot of the page on cartoonist Jeph Jacques or Bill Thayer’s experience cited above. Whether this policy-as-enforced is really the best reading of Wikipedia’s policy-as-written I leave to the lawyers.