Interesting read, and it reminds me of a really annoying experience I once had with one of the StackExchange sites.
There was a question about Emacs, and the answer was a chunk of elisp that was wrong - one of the closing parens was in the wrong place, so it caused an error. OK, easy fix! But I don't have enough karma to edit answers on StackExchange and have them submit cleanly, so I submitted an 'edit suggestion' or whatever they're called.
Unfortunately edit suggestions can't be submitted unless they have at least five characters of difference - seems like a fair rule to prevent tiny changes but five characters of difference can be a world of difference in elisp.
So I pad out the edit suggestion with a comment that added some value, only to have the suggestion rejected by a user who freely admitted to having "no emacs experience" because he "didn't think it looked valuable".
--
It also reminds me of a conversation I had with a professor of mathematics that taught one of my classes. He said he started and stopped editing Wikipedia the same day - when one of his changes to a graph theory article was reverted by someone who was "a computer repair technician" because of some Wiki-rule.
It's very possible that he did break the rule and his edit should've been undone, but at the same time he's the editor that Wikipedia desperately needs to attract.
--
Sites like this have Harrison Bergeron's Handicapper General ensuring that all contributions are seen as "equal", so it's not at all surprising that the article identifies one of the more comprehensive Wikipedia articles as "List of Pokemon".
Wikipedia needs to figure out how to say that the guy who wrote a paper with Erdös gets more leeway on mathematical articles than a guy who repairs computers all day.
I bet that your professor tried to write content from his own knowledge rather than strictly from published sources.
The problem is that anyone can claim to be a math professor with expertise on graph theory. For this reason, everyone - even experts - need to cite sources in order to add content to Wikipedia.
Physicist here: I stopped making contributions to Wikipedia after a few edits and reversions, I stopped too. Furthermore, the danger that something I'd put a lot of time into could be replaced meant that it's not worth sinking a lot of time into it. The citation policy makes a ton of pragmatic sense, and I understand why it's there. It's an unfriendly way to meet the Wikipedia community. (an alternative way to address a "citation needed" concern is to presume that the editor is correct, and go looking for the citation)
In addition, I'm not aware of any way to be compensated, even by citation, for putting time and care into Wikipedia. As academics, we give a lot away for free (want to know some of our most-clever trade secrets? Just read our papers, where we detail everything we do so you can replicate it), but our compensation ultimately comes from what we can show others as our contribution to society. To anyone considering me for a job, promotion, or tenure, a well-written introductory article on precision tests of gravitation isn't worth a hundredth of one precision test of gravity published in Physical Review Letters.
For a single contributor, Wikipedia is ephemeral, but the academic literature is effectively permanent.
(an alternative way to address a "citation needed" concern is to presume that the editor is correct, and go looking for the citation)
You mean, "hey, I wrote it, you go look it up if you don't believe me?" Down that path lies madness. And Fox News.
Academics aren't supposed to have "trade secrets" by definition, because that impedes the teacher and researcher's motivation. Historically, anyway, by my understanding.
I don't think that's what is meant by "go looking for the citation."
I read it more as "by being an expert in this subject I know this is true; by being an expert in this subject I also probably don't have time to find the journal that has the article in question, cite it per Wikipedia's rules, and put it in the article."
Frankly that sounds like a great task for editors that aren't subject experts. Maybe the expert can say "this is missing knowledge, this is where I think a citation should be or what to look for" and the non-expert can do that.
If you do remember a journal article, it's perfectly fine to put in an incomplete citation with a note, and someone will fix it up. <ref>Einstein's paper on foo (FIXME: fill out this citation)</ref> or something is typically sufficient for someone to track it down. It's not sufficient to just say that you know it's true, though, for good reason; a lot of "common knowledge", even among scientists, is incorrectly or imprecisely remembered. Typically we are most reliable in our specific sub-specialty, where we actually will have no trouble coming up with sources off the top of our heads (in my very specific sub-sub-specialty, I could rattle off sources for almost anything without even needing to look them up). Outside our sub-specialty, we may remember some things but not 100% reliably, and often based on obsolete or incomplete knowledge (e.g. a course from university years ago) that someone familiar with the literature would be able to do a better job of covering. Hence the importance of referring directly to sources, not only to cite them, but to make sure the information is actually in those sources precisely as remembered. To me, "I know X is true, but I don't know where you can verify that" is a bit of a red flag, typical of e.g. something someone vaguely remembers from med-school or undergrad, rather than something they really have a solid handle on.
In any case, as an academic myself, I find it weird that we, of all people, would find ourselves "too busy to reference" or something. That's an attitude one finds among students asked to write papers, but academics do not have problems peppering supporting references all over the place; it's basically second nature. Some academics even get in arguments with book publishers because the publisher always want to remove the "clutter" of too many citations and footnotes that intimidate the general reader...
I read it more as "by being an expert in this subject I know this is true; by being an expert in this subject I also probably don't have time to find the journal that has the article in question, cite it per Wikipedia's rules, and put it in the article."
Yes, I think you'll find that's explicitly what I'm advocating.
Maybe there's some sort of verification. Maybe experts get limited to their subject area. Maybe we just have to (gasp) assume good faith on the part of the editor and trust that they are experts in what they claim.
But I'd happily accept the edits of five "experts" if it means that someone that published with Erdos sticks around to edit.
I completely understand if that's not what Wikipedia wants. But I also think it's the wrong decision.
Wikipedia could add a category of academic editors whose identities have been verified in some way with their institutions, and who meet some definition of subject area expert. These editors would be given a presumption of innocence by other editors, so their changes should be researched for citations, rather than reverted.
One of the great things about Wikipedia is the ability to follow citations and read the original material.
To allow "expert opinion" without opening the floodgates to everyone with an opinion would require some kind of validation, which is probably as much, or more, work as citing sources.
> Frankly that sounds like a great task for editors that aren't subject experts.
I disagree. Hunting citations doesn't take specialized knowledge, but it does take time. I doubt that the ratio of editors to article churn would sustain them being responsible for hunting citations.
I think that's really the core of most issues with wikipedia: neither side has the time. Individual contributors don't have the dedicated time to comply with all quality-control policies themselves, and editors are spread too thin to provide any feedback beyond quoting chapter and verse.
Sure, and if the claim is actually controversial (someone says "I don't think that's true) we need higher standards. But I see stuff reverted without anyone willing to say "that sounds wrong".
And we don't have trade secrets, that's why ISL followed it up with this:
> want to know some of our most-clever trade secrets? Just read our papers, where we detail everything we do so you can replicate it
ISL was pointing out that many other professions have a lot of field specific knowledge with which they can charge people money to access. Academics print all of the information for basically free. So we already give it away for free, we don't have the time or patience to give it away free to somewhere that's not going to keep our work.
I think you have an interesting point about recognition for your work. It seems to me that Wikipedia's back end is ferociously clunky and outdated, there is no easy way to show all contributions made by a specific user, for example. If you compare this to a system like GitHub or one like StackOverflow, the deficiencies are obvious. Us programmers can easily point to our contributions to open source projects or our general knowledge through pointing at our profiles on those sites. A software overhaul to make Wikipedia more social, easier to edit would also go a long way.
Please come back and make suggestions on Talk pages, rather than directly edit. There, you can actively discuss, cite sources (your own and hopefully others), and make real contributions which will last. Experts have a special kind of conflict of interest: their possible bias when directly editing. Discussing will moderate and ameliorate that bias, before it gets into the article. This approach will help as science advances and understanding deepens and changes: recent advances in physics prove this. We need your help; just not in direct editing. Your contributions in Talk will be real and welcomed (and maybe disputed, but never deleted) by our more experienced, sane, editors.
(typo corrected)
Frankly, I'd rather not have the Talk page or the "editors", and a lot more actual data input.
The barrier for entry to writing articles has to be way, way lower than it is; it shouldn't require a negotiation with a bunch of self-important petty bureaucrats.
What those self-important guys (they're mostly guys) miss is that the best is the enemy of the good. Their misguided approach to quality actively prevents improvement.
Outside the high-edit-count word-shufflers, I'd classify contributions to WP in four buckets: (a) substantial contributors who know a lot about a little and are able to significantly add or improve content; (b) trivial fixers, who fix grammar / spelling mistakes, or otherwise make tiny casual edits; (c) mindless vandals, who work some adolescent scrawl somewhere in the text, with varying degrees of visibility; and (d) the really sneaky stuff; well-written fakes, PR companies, reputation massaging, etc.
The guys in (a) need to be preserved at all costs. I think the editors are concerned about (d), but the trouble is that they're doing it in a rules-bound way, but because knowing the rules and right WP:INCANTATIONS is a source of power, the rules become a thing in themselves, rather than a last resort for eliminating (d).
I'd much rather have a more complete, comprehensive WP than one that's even 90% correct.
New contributions to established articles should often be discussed on the talk page because of (a) though, and the fact that people sometimes just come in and start making major changes without looking at the work that's already been done is one thing that actively drives experts away. After a few history experts discuss the conflicting sources and an article reaches a semi-stable state after considerable discussion and additional research, someone will come in and make major changes to the article without reading any of that, usually based on an incomplete-at-best understanding of the historical sources, and sometimes reintroducing popular myths that had been previously removed after library research turned up good sources on the subject. Sometimes there are legitimate objections and an article should be changed. But also, sometimes, certain questions or disputes or uncertainties have already been discussed in previous versions of the article (often by historians who work in this field!), and it's worth understanding the work that's already been done to get the article to its present state, before rehashing the same issues.
A number of historians have given up on Wikipedia because it's too easy for people to edit articles without discussion. An expert will spend a bunch of time working out a solid, cited article that integrates the best current understanding, often in discussion with other people, then come back 6 months later and find their work has basically bitrotted. In which case, why bother doing the hard work of hashing out a good article in the first place? Some academics therefore prefer a much more bureaucratic encyclopedia model, where all changes must first be proposed and then vetted by an expert in the subject who "owns" the article; Citizendium and Scholarpedia are two attempts to build that model of encyclopedia.
In that case, would it not make sense to have metadata associated with a particular section, detailing the final consensus and how it was reached? Kind of like a comment describing a non-obvious section of code. Otherwise you have the equivalent of telling someone "Before you modify this code, look through the entire git history to see why it's the way it is today".
Yeah, there have been some discussions of that, but nobody has put together a MediaWiki plugin to do it afaik. Better annotation of regions in general is a longstanding wishlist, so discussions or even citations (or requests for citations) could be attached to a region of text. But then there are problems like how to handle annotations and arbitrary editing/splitting/moving of regions.
As a low-tech solution, you can actually put HTML comments in the source code, which people occasionally do, but it doesn't seem to be a very well-known option. <!-- Do NOT change birth date to XX/XX/XXXX, see talk page --> kind of things sometimes appear in the source. I haven't seen them used for longer discussions, though, just one-line "hey, watch out before you do X" things.
On contentious articles with a long history of debate people will sometimes write a summary on the talk page, so you don't have to read through the whole history of the discussion. On most articles, though, the talk page isn't huge, so I find it easy to glance at before making changes.
If you have a problem with experts disagreeing about facts, arguing over them, reaching a consensus, and then other "experts" starting up the argument again - it really sounds like you're doing a poor job of communicating the researched consensus in the relevant article.
When that sort of stuff happens, I think you should make sure the incorrect information is referenced in the article as a misconception, or just have the correct information associated with citations. False information should only be erased if it is not "in the wild", so to speak.
First, I think Wikipedia is great. And I don't want to imply that the Talk pages are bad. But, I really find the Talk pages to be the most "inside baseball," almost inane, and just impenetrable thing on the Web. Might just be me.
And your contributions can be even more easily ignored when they're made on Talk pages. If you provide useful information or links, you can expect it to be ignored. I ran a little experiment http://www.gwern.net/In%20Defense%20Of%20Inclusionism#sins-o... where I provided high-quality RSs/ELs, and the uptake rate after a period of years ranged from 3-8%.
If that's true, then incumbent Wikipedians need to take it upon themselves to communicate that to newcomers directly, and not in a random comment in Hacker News.
Probably the easiest way to do that: If someone makes a substantive edit that isn't obviously vandalism, but doesn't quite meet Wikipedia's standards, then don't just undo it wholesale. Instead, temporarily pull it off the main page and into a back-end discussion. And then immediately contact the author with an explanation of what's going on and what they can do to get the changes reinstated.
Well you can keep his edit and send the guy a polite message that he ought to cite his sources soon or you'll remove what he added. I mean, at worst you'll just have an article without a citation for a short while, without driving people away. What you are doing right now is this:
Over my years of editing Wikipedia off and on, I've come to think this is mostly a bad excuse. The issue is that very few articles meet Wikipedia's real standards for citations. Yet new content provided by experts is often challenged for lacking citations, perhaps because of a sort of status quo bias.
What needs to happen is that there should be collaboration between Wikipedia insiders who know the Wikipedia system and subject matter experts who have valuable information to add.
Worse, the demand for citations is often just absurd. There's a big difference between "this is a controversial claim that might need citations" and "no one has yet bothered to look in a basic textbook for citations."
There are real reasons why experts can't just come in and say "do it my way" but the response has to be geared towards saying "how do we take advantage of your knowledge without forcing you to learn arcane policies?"
Update: As for identifying experts, that's not altogether obvious. But if an article is an absolute mess, there should be a low bar for cleanup. If the article is in decent shape, then presumably someone is running around with a bit of subject matter knowledge, and can coordinate. Think about any area you know something about: how quickly can you distinguish someone who generally is on target from someone who doesn't have a clue?
I was thinking about this the other day and I came to think that Wikipedia's rule about citations is this: Wikipedia does not want to be a source of information. It wants to be a collection/aggregation of other sources of information, and those other sources have to be accessible to others for all time (i.e. not a human being but a book written by a human being).
What I mean to say is that Wikipedia's system is one that does not/will not give article editors credit for original content; it treats them as just collectors and explainers of original content. When the explainers start writing stuff that has no source, they then disappear and leave Wikipedia holding the content which is now unverifiable. And that's just not how Jimmy wanted it to work I guess.
When I look at it that way, the policy doesn't seem that ridiculous. It may be inconvenient, but that's like saying a stack data structure is inconvenient because you can't remove things from the bottom. It's just the way it is; it has advantages and disadvantages.
I agree that Wikipedia's goal is to be the collection of things that we know that we know. So we do need citations. But when an article is wrong and doesn't already have thorough citations for every claim, should we demand citations for every new contribution?
My point is really about process: when someone comes to us with a bit of knowledge, but no citations and limited understanding of the Wikipedia process, how do we react?
The model whereanyone can edit and see their changes live has real benefits and costs. It makes it easy to make your first contribution, but it also means that if an editor doesn't like it, they just revert it. Sometimes, maybe changes should go in a queue or something.
I come to Wikipedia, and dump a ton of graph theory on some page. An editor says "look, this isn't quite how it works, but it looks like you're trying to improve the page. Is this stuff you have citations for? Do you know what textbooks would cover it?"
Maybe my edits don't go live immediately, but it's better than the current situation, where they just get reverted by some guy spouting WP:STQ!
(That's the grain of truth the comment down below that Wikipedia needs to be more like git).
> But when an article is wrong and doesn't already have thorough citations for every claim, should we demand citations for every new contribution?
Ideally, you'd go through the article. You'd remove anything that is "wrong"[1], and find citations for everything else. Then, when someone wants to add anything, you can find a cite for it, or ask them to find a cite for it, and remove it if there is no cite.
This demonstrates why WP can be horrible - anyone doing this to any article would soon find themselves skewered in horrible WP processes. Good cites are mocked as POV pushing, hopeless cites are forced in by people with more time than you.
> I come to Wikipedia, and dump a ton of graph theory on some page. An editor says "look, this isn't quite how it works, but it looks like you're trying to improve the page. Is this stuff you have citations for? Do you know what textbooks would cover it?
Sometimes that's how it works. There are people on WP who are great at helping that style of new editor; finding them mentors or whatnot to help them with the process. it's hit and miss, sometimes it fails badly.
> where they just get reverted by some guy spouting WP:STQ!
WP has made some effort to prevent the over active 14 year old using tools to auto-revert hundreds (thousands) of edits per day.
I agree that many articles lack sufficient reliable sources. Only college or postgrad textbooks are considered reliable sources, but not great ones, unless written by noted and/or frequently cited authors.
In my opinion, experts should be sincerely welcomed to kibbitz on talk pages - not directly in editing articles - with a strong invitation to cite sources other than themselves. This approach, optimistically, will keep bias on the Talk page, and out of articles, and give experts a substantial real and perceived voice, currently lacking.
Yeah, but the talk pages are basically just sewers for a lot of articles. I look at a talk page and my thought is "god, no one will ever read what I write here".
What you can do is first start a section on the talk page about the topic and which sentences/sections could be changed. Then go ahead and make the intended edit on the article and then link to the section on the talk page in your edit comment. Then any editor who may potentially revert or have an issue with the edit will see that link and can go directly to that talk section that you started.
The markup used for the linking in the edit comment is the same as the wikitext markup. For example, an edit comment like this:
... see talk page for discussion.
would be marked up:
... [[Talk:Article name#Section heading|see talk page]] for discussion.
This might help start discussions about low-traffic article topics that would otherwise languish at the bottom of the talk page for weeks if not months.
Oh, that's another association for sewer that I wasn't thinking of. I didn't actually mean that the comments are crap or nasty, though they sometimes are. I just meant that it's a big backlog of comments that people would probably not read through.
The set of experts I'd like to invite to Talk page discussions would be published, widely cited authors, including professors, engineers, scientists, and non-degreed practitioners with a good reputation in their field. Publications can be academic, professional, and/or popular, IMHO. Peer review, of course, helps. Longevity helps. Professors who have published, and been cited, would certainly become seen to be established in their field.
I bet he was too. But personally I don't care, and in my opinion neither should Wikipedia.
In my opinion it'd be better to risk a little inaccuracy if it meant attracting experts to edit articles. But as cldr notes above, it's possible that's not the goal of Wikipedia - which is completely fair, but makes me wish some hybrid of NuPedia and Wikipedia existed.
I gave up StackExchange since day 1, I found no difference between being registered or not, any useful action is blocked because I'm a new user, and I would rather stay away than spend time figuring out my way through a hostile site like that.
If you care, or for others who haven't yet approached Stack Exchange sites [1] in the wrong way and found themselves put off as a result, the way to start building reputation is by answering questions, not just accurately but as extensively and generally as you can manage, with bonus points for considering edge and corner cases. Play to your strengths in terms of domain knowledge, especially in relatively obscure domains of which you have extensive knowledge, and don't get discouraged if you are not immediately lauded; SE pages tend to rank very highly in Google searches for topics to which they have relevance, which means that if your answers are sufficiently valuable, they will be found and upvoted over time; in terms of reputation, which SE makes the essential currency of being able to do things such as edit questions and answers directly, this 'long tail' effect is considerably more valuable in the long term than getting your answers accepted. (That green check mark is worth 25 reputation points, once per answer; each upvote is worth ten, as often as someone with an account finds your answer of use. Answering the specific question in a quick and narrow fashion is more likely to get you the check mark; answering extensively and in detail, as I describe here, will certainly get you upvotes over time, unless no one but you and the asker cares about the subject anyway. Which buys you more rep? You do the math.)
Is it necessarily worth playing the Stack Exchange game? Certainly not, especially as the sites have grown to an extent which strains the moderators' ability to manage them -- Super User, I'm looking at you in particular, and the flood of undifferentiated garbage questions which has increasingly occupied your front page since I created my account a year ago, and Stack Overflow, you're hardly better, thanks to the endless zombie hordes of idiot PHP hacks who ask the same stupid question twenty times rather than try a single Google search which would certainly turn up the obvious, simple answer. Worse, in fact, as the original developer admits [2], the Stack Exchange model is explicitly designed to be addictive, and he and Spolsky nailed it; if you're not in favor of time sinks, Stack Exchange is probably not for you. But if you're a domain expert who finds interesting questions appealing, or even better someone who finds that the best way of gaining knowledge of a given domain is to work out how to answer other people's questions about it, you might do well to try out a Stack Exchange site.
[1] Except for English Language and Usage, than which you will never find a more wretched hive of snark and pedantry; unless you're a tenured professor of pointless linguistics, your time will be better spent on almost any other activity, be it Wikipedia editing or toenail bonsai.
The problem with StackExchange is that it's turned into the Eve Online of technology forums. It's fun for the few people who managed to get in early or got lucky and asked a question or posted an answer with lots of upvotes. And, simultaneously, it has a forbidding learning curve for everyone else.
I was involved in StackOverflow when it was still fairly young (before the whole StackExchange concept had taken off). I asked questions, I answered questions, I got up to ~1k points and then I stopped going for a while. When I returned, I felt distinctly unwelcome. The social mores had shifted to such a degree, I felt like I was coming in as a new user rather than one with some experience at the site. The experience was very off-putting, and I consider a large part of the reason why I haven't made any contributions to StackOverflow in a long while.
My point exactly, or if I find a useful answer in a completely obscure question and I want to upvote the person that answered it because she got too few votes (xx points required).
Like I said, I got frustrated because I couldn't do anything useful, and I would rather stay away than be hunted down by a SE Inquisitor.
Sorry to hear you had that trouble on Stack Overflow -- I assume it was SO, anyway; most of the Emacs questions tend to live there.
I doubt you care at this point, but for others who might, it's generally a better alternative for a new user to comment with the correction, than to try to edit; comments are globally visible no matter who makes them, and a comment pointing out an error like that one is very likely to result in its being edited into the answer by someone who does have the reputation score necessary to do so. Even if it doesn't, someone who tries to use code from the answer, and finds it does not work, will be able to read the commentary and find out why not and how to fix it.
As for the edit reject reason -- well, all I can say is that idiots get everywhere, Stack Exchange sites not excepted. For whatever it's worth to anyone, I've found that sort of obstreperous idiocy much less common on most SE sites (again, English Language & Usage excepted!) than in, for example, Wikipedia.
Comments are only available at 50 reputation points (or some number requiring a large number of previous reputation building actions), which a new user wouldn't have.
Are they? It's been long enough since I started that I'd forgotten. Still, 50 points is 5 upvotes, which is not difficult to attain quite quickly -- I earned >400 rep in my first two days with an account on Super User, and while that's not typical and took some effort on my part to achieve, a mere 50 points shouldn't be all that hard to manage by comparison.
This brilliant professor of yours should not have a problem following a few simple rules that a repair technician can and everyone else abides to.
> he's the editor that Wikipedia desperately needs to attract
If those editors require special care and exemption from the rules on the account of their fancy credentials, they're certainly not the type for Wikipedia.
One of those examples is ISL the physicist here in the comments. He complains that his contribution got edited by someone else. And that he doesn't get compensated for his hard work on the wiki. I'm sure he's a smart fella, but this attitude is not compatible with the Wikipedia model.
McDonalds is surely turning aways scads of customers by not providing their food for free.
Certainly there are many things Wikipedia could do to be more welcoming. But not permitting editing by others and/or paying people to edit are not within the Wikipedia model any more than McDonalds giving away all their food would be.
There is nothing wrong with being polite but firm about what something is and what it isn't.
In my view, a large part of the problem is that many of the rules or the way they're enforced borders on the idiotic.
Take [citation needed] as a case in point. It's aggressively and mindlessly waved all over the place, irrespective of pre-existing content in wikipedia itself and obvious factual information.
Say you contribute one or two sentence summary of related pages -- complete with links. Or worse, a sentence or two that expands on an existing statement by merely highlighting the obvious assumptions or implications. Such edits have good odds of being greeted with the dreadful [citation needed] or reversed outright by some random rules zealot.
Adding insult to injury, the latter seem perfectly content with sources such as The Onion and slimy news sites. The thing that really matters is that it must be written somewhere else. It's then up to random editors to decide whether that somewhere else is credible or not.
Admittedly, would-be editors can jump back in to defend their edit, appeal, etc. and even expand on it with the requested citation. But this seldom occurs, because it leaves such a sour taste in their mouth that they grunt in bemusement, scoff the whole thing as hogwash, and move on with their life.
It's sad that its misused like that. When I created it, I just wanted the project to be more reliable, via sourcing. It was never meant as a weapon of war!
This is by far wikipedia's biggest weakness. The people with the power at too disconnect from what matters. Even worse, they have a false sense of their own importance. And wikipedia's reference guidelines are silly at best.
Wikipedia wasn't designed to scale to what it is now, and a lot of the erroneous assumptions baked into the design are causing growth of the site to become more and more unsustainable.
I think you could insert any large system (Facebook, bigcorps, governments, etc.) in the place of Wikipedia in your comment and it would be accurate. What can be done about it?
It's said that the best way to be a primary source for wikipedia is to write a webpage that posits the change you want to make to a WP article, then cite that page when you make the change.
As a world expert on some topics (Machinima filmmaking) for which the primary source of some relevant information resides largely in my head, I was wondering if this was the best way to approach Wikipedia edits. Thanks for the info: I may actually make some contributions soon!
If what you know has been published in a way which meets enwp.org/WP:Reliable_sources then it can be used in articles. Unpublished material can't be used, because it can't be independently verified by anyone - see enwp.org/WP:V
Self-published material might be usable under certain circumstances, if you're already an established expert in your field (per independent reviews, or citations by others) - see enwp.org/WP:SPS
I thought Wikipedia wanted secondary sources (i.e. newspapers, journals, textbooks, biographies) that have reported on primary sources (individual papers, original research).
I also thought it was frowned upon to cite your own research.
You are correct. WP still does want secondary sources, exactly as you state: http://enwp.org/WP:V and WP:IRS Independent reliable sources, and does not want WP:OR. IMHO it's ok to discuss one's own work on an article Talk page, in the light of a future improvement to the article, as in, "I've just finished this, it's in peer review now, it's slated for publication on (future date), and it's relevant to update/correct (section). I'll come back and notify y'all if it receives commentary by others elsewhere."
I think Wikipedia must evolve like old search engines evolved to using a pagerank.
Obviously I don't know how to solve this issue but at first editors must be evaluated by the audience and their scores accessible. All the other Wikipedia processes are bureaucratic .
There was a question about Emacs, and the answer was a chunk of elisp that was wrong - one of the closing parens was in the wrong place, so it caused an error. OK, easy fix! But I don't have enough karma to edit answers on StackExchange and have them submit cleanly, so I submitted an 'edit suggestion' or whatever they're called.
Unfortunately edit suggestions can't be submitted unless they have at least five characters of difference - seems like a fair rule to prevent tiny changes but five characters of difference can be a world of difference in elisp.
So I pad out the edit suggestion with a comment that added some value, only to have the suggestion rejected by a user who freely admitted to having "no emacs experience" because he "didn't think it looked valuable".
--
It also reminds me of a conversation I had with a professor of mathematics that taught one of my classes. He said he started and stopped editing Wikipedia the same day - when one of his changes to a graph theory article was reverted by someone who was "a computer repair technician" because of some Wiki-rule.
It's very possible that he did break the rule and his edit should've been undone, but at the same time he's the editor that Wikipedia desperately needs to attract.
--
Sites like this have Harrison Bergeron's Handicapper General ensuring that all contributions are seen as "equal", so it's not at all surprising that the article identifies one of the more comprehensive Wikipedia articles as "List of Pokemon".
Wikipedia needs to figure out how to say that the guy who wrote a paper with Erdös gets more leeway on mathematical articles than a guy who repairs computers all day.