Sources don't have to be free, it's just prefered. That article doesn't assert notability, though, so it should probably be speedied.

On 03/01/07, The Cunctator <cunctator(a)gmail.com> wrote: I've always thought the underlying justification for unnecessary deletionism (the deleting of articles which are factually accurate and verifiable) is elitism. IMO, it detracts from Wikipedia. -- Oldak Quill (oldakquill(a)gmail.com)

On 1/3/07, Bogdan Giusca <liste(a)dapyx.com> wrote: Right.

On 1/3/07, Gregory Maxwell <gmaxwell(a)gmail.com> wrote: Where did nonverifiability come into the discussion? "We cannot survive"? Is someone predicting the imminent death of Wikipedia? What problems are being ignored? This is a distinct issue from inclusion vs deletion. Who is expressing hate for anyone? On 1/3/07, Bogdan Giusca <liste(a)dapyx.com> wrote:

Because we have a policy that says we delete articles that don't assert notability. If you want to change the policy, then you know what you have to do.

Ok, so I generalised, I'm sorry. My point stands, though, as we are talking about "web content". CSD is policy and it says that articles about websites that don't assert notability (or "importance and significance", but notability is close enough) can be deleted. True, it doesn't say "must be deleted", but that's just semantics when the deletion is allowed to be completely unilateral. It just takes one admin to want it gone, and it's gone. If people want to keep such articles, they need to get a concensus to change A7.

On 1/3/07, Ken Arromdee <arromdee(a)rahul.net> wrote: And why is that? Remember, people learn to write Wikipedia by reading Wikipedia. If we're going to relax our content standards substantially for one area, people are going to carry the lessons they learn from reading that area into the rest of Wikipedia. If something can't be sourced up to our standards, we can do without it. -- Robth (http://en.wikipedia.org/wiki/User:Robth)

On 1/3/07, Ray Saintonge <saintonge(a)telus.net> wrote: It's a hard theory to provide specific evidence for, seeing as it does involve trying to get inside people's heads, which is all but impossible on the internet. What I do think we can reasonably state, though, is as follows: 1. People learn more about how to write Wikipedia from reading articles than they do from reading style guides. I know this has been the case with me, and I haven't met anyone with whom it isn't. We have an ungodly amount of guideline material relating to what articles should look like, and anyone who attempted to read it all before sitting down to begin writing for Wikipedia would get bored and give up before they ever started typing. Style guidelines are all well and good, but we have to acknowledge that, at the end of the day, the drive-by contributors who account for most of our material are, in the best case scenario, going to write something that looks like other articles they have read on Wikipedia. The better the average article is, the better the average passer-by contribution is likely to be. Think how great it would be if just 1 out of every 10 college kids who make a drive-by contribution to Wikipedia went and got a book from the library, checked their facts, and cited their sources when they wrote. That isn't impossible, but it would require that our across-the-board quality be high enough that getting the book would seem like the natural way to contribute to Wikipedia. Quality begets quality. 2. You can't quarantine topics from each other. Now I'm not arguing that people are going to read poorly sourced webcomic articles and then immediately go write articles on medieval Scandinavian literature sourced from the same blogs. You refer to people who are "only interested in serious topics", but people do read and write about more than one topic apiece on Wikipedia. And there is a startling amount of really shitty content about serious academic topics on the web, waiting for people who have learned to look to google for their sources to come snap it up. I don't use web sources when I'm writing, but every now and then I google the topic I'm working on and am blown away by the sheer quantity of incorrect information there is out there. If people observe that "Some Internet Guy said it" is accepted as a reasonable source for large portions of our site, they're going to go look and see what Some Internet Guy has to say about medieval Scandinavian literature when they decide to help out Wikipedia by writing an article about this cool book they just heard about. And the article they write will be, as a result, bad and inaccurate. (Really, google an academic topic you're familiar with, and imagine just how atrocious an article based on the sites that come up would be. I suspect that history is the worst, as some basic storytelling impulse inspires people to write breathless semi-informed narratives about stuff they took a class on one time, but it isn't alone.) 3. We don't have the manpower to contain the spillover. This is my problem with the argument that we can allow Some Internet Guy to serve as our source for articles about stuff that only Some Internet Guy cares enough about to write about, but then, through rigourous enforcement of our standards in other topics, ensure that only reliable sources are accepted for most subjects. Now this might work for subjects like Israel-Palestine, where both the IDF and Hamas have full-time personnel vetting every single edit (or have we not reached that point quite yet?), but it won't work for the vast majority of topics, in which most articles are monitored loosely or not at all, and any edit that isn't vandalism tends to stick. Remember, source quality and style guidelines are invoked only in those rare cases where two people find themselves working on the same article at the same time; editing in a fairly popular academic subject area, I have seen such simultaneous editing on only two or three occasions (outside of the FA or GA processes) in my year here. We don't have the resources to maintain the kind of line that seems to be envisioned in many people's comments on this topic. I suppose that what you take away from this depends on your perspective. I'm sure there are people who would argue, given these three points, that we should do our best to contain this spillover and accept what we can't contain as a necessary byproduct of having these articles. I don't reach that conclusion. By including these difficult-to-source subjects we make available information, to our present internet audience, on topics of short-term interest (let's not kid ourselves about that) that are already covered in spades by the rest of the internet. This is a good thing. If, however, by doing so we impede the development of content that will be of use to a much larger potential long term and offline audience, then forget it. Including information on websites, fads, webcomics and the like is a nice thing to do, if it can be done without spilling over and worsening our coverage of stuff that people are still going to care about 10, 20, and 50 years from now. But if covering those subjects involves lowering our standards in a way that adversely effects the average quality of incoming contributions, and the standard or writing on other subjects--and I believe that it does--then forget it. This was long. Sorry about that. -- Robth (http://en.wikipedia.org/wiki/User:Robth)

On 1/4/07, The Cunctator <cunctator(a)gmail.com> wrote: Bad, inaccurate sources=bad, inaccurate article. To pick an example I'm familiar with, here's are the first five non-Wikipedia google results for the first major article I ever did for Wikipedia, [[Epaminondas]] (filtering out totally off-topic or near-duplicate results). http://www.fordham.edu/HALSALL/ANCIENT/nepos-epam.html -Ancient primary source of dubious validity. Usable only with input from reliable modern sources. http://www.in2greece.com/english/historymyth/history/ancient/epaminondas.htm -Utterly atrocious page written by Some Internet Guy, containing numerous inaccuracies. http://www.mainlesson.com/display.php?author=haaren&book=greece&sto… -Dramatized story of the man's life. Glosses over critical aspects, quite inaccurate on some points. http://www.nndb.com/people/812/000095527/ -Best of the lot, but contains numerous inaccuracies and omissions. http://www.lbdb.com/TMDisplayLeader.cfm?PID=5744 -Short You could write an article from these sources. It would omit numerous important points, and would probably also include numerous untruths. Such an article would be, in my opinion, worse than nothing. And I have seen sites like these used as the basis history articles; here's one of my favorites: http://en.wikipedia.org/w/index.php?title=Methuselah&oldid=71757344 -Article bloated with dubious speculations drawn from the internet (not all of them directly cited, but google shall reveal for those that aren't) and of course, the requisite In Popular Culture section. And actually, looking back, my rather timid cleanup probably wasn't sufficient. That's obviously a pretty susceptible article, but it isn't the only example, just the one that jumped to mind for me, and I can't convince myself that the belief (inherent in the edits that produced that page) that Wikipedia is a place to put information you found on the internet doesn't in some part stem from the fact that we have a number of articles that are, well, a place to put stuff you found on the internet. -- Robth (http://en.wikipedia.org/wiki/User:Robth)

On 1/4/07, Robth <robth1(a)gmail.com> wrote: How inaccurate the online sources are depends on the subject matter, however. Generally internet sources are not much worse than the popular press when it comes to accuracy; such an article, pre-Internet, would likely pick up the same kind of errors from the popular press if they were used as the source. For some subject areas, the academic works are published online and available. In others, there are no academic works to use, and sometimes the most authoritative works are online. -Matt

On 1/6/07, Ray Saintonge &lt;saintonge(a)telus.net&gt; wrote: I'm wasn't talking here about writing style, which is something that everyone has independently of Wikipedia, but referencing style. Very few people come to the project with established preferences for how to research and reference encyclopedia articles. And while it's true that you can offer suggestions for improvement for people who come to stay, the same is not true of the short-term contributors who account for most of our content. Our one chance to reach many of our contributors is through the articles they read before contributing, and if the majority of those articles are unreferenced or poorly referenced they are going to (as I did and I suspect many others here did at first) start by writing unreferenced or poorly referenced articles; if they only work on one article, that may well be all they write. This is an example of a view of the Wikipedia system that I find very unconvincing. First, the notion that articles organically develop and improve over time, with each stage of the article serving as the building block for the next, is not supported by examining articles that expand and improve substantially from their creation; generally, such expansion is a result of a single editor replacing (as opposed to slowly expanding upon) the existing content (an interesting essay on a subject related to this can be found at <http://en.wikipedia.org/wiki/User:Matt_Britt/Don%27t_just_do_whatever>). While there is certainly a value to having something, whatever it is, at a given location (breadth of coverage is a sort of quality in and of itself), we shouldn't overstate the role that existing weak articles play in developing strong articles to replace themselves. Furthermore, I think the almost dialectical view, evident in your comment above, that Wikipedia is headed inexorably in the correct direction is an incorrect and dangerous one. While the project has already achieved a great deal, it is still quite possible that it could end up as something much less than it has the potential to be, that it could become largely a compilation of the collected wisdom of the internet instead of a true collection of human knowledge. So long as the volume of content from passer-by contributors continues to dwarf that from regulars (i.e. for the foreseeable future), the quality of referencing in contributions from new users is going to be the greatest factor in determining the direction we move in in this regard. Taking a lassez faire approach to article quality (and specifically, reference quality) will ensure that our average article remains rather poorly referenced, and strikes me as a sure way to at best delay the development of a higher quality encyclopedia, and at worst lastingly sidetrack the project. -- Robth (http://en.wikipedia.org/wiki/User:Robth)

Wednesday, January 3, 2007, 7:05:16 PM, Ken wrote: Maybe I spent too much time editing history articles, where the source credibility matters, but accepting blog/forum/usenet posts as valid sources would be a great mistake, IMO.

On 03/01/07, Bogdan Giusca &lt;liste(a)dapyx.com&gt; wrote: A one-size-fits-all approach to what constitutes a reliable source is exceedingly foolish. The sourcing needs to match the subject matter. (The news media is not "reliable". It's "easily verified", which is not the same thing at all. Many blogs are a hell of a lot more reliable for factual information than many newspapers.) - d.

Steve Block wrote: As far as I'm aware, the OED cites such sources as examples of words in use. In effect the OED is saying "this word was used in this document on this date", and citing the document itself to show this. This is not the same thing as providing a citation for a factual claim made in an article. -Gurch

Steve Block wrote: The difference is that OED editors are experts exercising professional judgement and putting their reputations on the line when they make their assertions. For instance, if you look closely, you'll see that they're very careful about claiming to know when a first usage occurred. I don't know how OED does business in the modern age, but I would expect they're screenshotting web pages and keeping them on file permanently, just as they once did for the scraps of paper that contributors sent in. WP *really* doesn't want much to do with primary sources; very few amateurs even know how to evaluate the trustworthiness of a primary, or how to reconcile multiple contradictory primaries, and we just embarass ourselves when we play at that. Unfortunately many WP editors are unaware of how much they don't know. Stan

Steve Block wrote: Well then, how do you verify? How do you know a primary source isn't flat-out lying? Experts write whole monographs on the massive extent of falsification and error in the primary sources for Roman history, for instance; other experts then use that analysis in constructing a plausible narrative of what happened. Our role is as glorified stenographers for all these experts; we summarize their results and withhold judgement on which version is "correct". That's why I didn't say "never use primaries"; I've used them myself. But they are tricky to use correctly, and if a large fraction of the community doesn't have a clue about scholarship, I'm not going to take that as a reason to give up and let WP turn into everything2 or fark or whatever. If a WP editor is opposed to the idea of doing scholarship, they should find a different website more amenable to their tastes; there are lots of popular wikis to choose from these days. I've been working on WP steadily for almost four years now, and if you look at my oldest edits, you'll see many of them (though not all!) including reputable secondary sources, and I did that because I read it on a policy page. So this isn't some kind of new thing. If all this really is "strange and varying criteria" to you, you really need to stop and consider whether WP is the project you want to be involved with. Stan

Steve Block wrote: Well, if the source is lying, then by definition it's not reliable. Your example is an object lesson in how newspapers are intermediate in reliability; they are usually better than Joe Random's blog, but not as good as a scholarly monograph that has had multiple layers of review spread over multiple years. Seeing the references to that article, I can see where you're coming from. I personally would be very reluctant to, for instance, use rootsweb to source a death date, because not only do we have the problem of identifying *which* person of a given name is meant, but genealogy sites include all the standard howlers, like descents traced to Julius Caesar, which is only plausible if you don't realize how utterly corrupt the primary sources are for the Dark Ages. The "California Death Records" link comes up blank, the oldpoetry.com link has a single unsourced line affecting to be written in the first person, post-mortem ("I lived from 1887-1972.") - kind of spooky actually. So some of your references illustrate well the reasons to be wary of primary sources. I bring these up not to try to disparage you, but because to me it's what is interesting about scholarship and Wikipedia. Which sources of information are good, which not so good, and why? If they're inconsistent, which is true, or are they both true if you interpret in a different way? I imagine that some day, if it hasn't already happened, a heated talk-page argument over some factual detail will inspire an expert to do some original research and then publish the findings - which we can then incorporate into the article originally in dispute. Stan

Steve Block wrote: Dangerously close to epistemology, but I'm not afraid. :-) I think we work with a fundamental underlying assumption that the facts are knowable, even if we don't know them at the moment. Otherwise, if somebody finds a hundred blogs stating that the sky is really green (perhaps as part of a prank), and runs to WP to add stuff about how the color of the sky is in dispute, we wouldn't have any basis for saying "dude, it was an April Fool's joke, we're not going add it". We've actually had a couple people in the past adding totally fictional events, sourced from websites that reported the fictional events as if they were real, and they got irate when we deleted their material, because they were convinced it was all real - or that was part of their prank, or possibly performance art, to this day I can't say for sure. So I exaggerated when I said we're just stenographers; we do have to do "research", if not "original research". That's why I said "personally reluctant", not "against policy". After thirty years of doing scholarship, I thought I was pretty savvy, but even so have been caught out a bunch of times while working on WP. Just recently I used a web article that looked good, and it cited four print references, but then we got a note from the webmaster of the site that he was pulling the article because it was inaccurate, and making claims not separated by those references. Ironically, the corresponding de: article is sourced from newspapers of the time, so now I'm planning a little quality time in the microfilm department of the library to try to sort it all out. Despite all our warnings, people do take WP at face value; I found it very sobering to see some of my early mistakes propagated onto websites all over the net. Maybe I'm the only one bothered by that. My argument is that web sources *may* have value, but people need to keep their hands on their wallets - we have a lot of editors getting their pockets picked and they don't even know it. Have you ever looked at the fine print in an unabridged OED? It's great stuff - those guys know about not only the English language proper, but every other human language, and all the historical context too. To them, a web usage is a raw data point, just like somebody's diary, or graffiti seen on a subway wall. You will never find an OED entry that cites a web page as an authority, it will always be in quoted form, as an example of observed usage. That is the crucial difference. Stan

Stan Shebs wrote: The assumption that the facts are knowable is necessary; without it we would be paralyzed. When we say they are known we often give our statements an absolutism that may not be warranted; we need to accept that some facts may never be known. There are degrees of knowability and balances of probabilities. The savvy reader needs to be able to weigh that information himself. The sky colour argument is not helpful. It is based on personal empiricism, and the more abstract process of how we define colours. In the context of the philosophy of science definitions are not falsifiable statements. In the example about adding fictional events as if they were real we need to assume good faith, but that gives us more worrisome results than using Ockham's razor to justify bad faith. Assuming bad faith makes the problem a lot simpler than it actually is; it allows us to say we _know_ what happened and to feel smug in that knowledge. Urban legends are often based on good faith interpretations of bad faith comments. The dangers of falling victim to such illusions and delusions are more obvious on the internet where the physical act of publishing anything is inexpensive, and the quantity of such publication is massive. We should not be lulled into a sense of false security about everything in print. "The Da Vinci Code" is a published book. I regret that we have many editors for whom "original research" is interpreted as "any research". Both require hard work, but that similarity does not make them synonymous. We have yet to face the full impact of a race of idiots solely dependent on the internet for all their information. It's all a matter of how effectively we and the end user can apply critical judgement. No matter how hard we try to avoit it we all sometimes get caught up using sources that later prove to be unreliable. No blame should attach to this, and the best we can hope for is a graceful exit. Genealogical sites are a hotbed of good-faith incompetence where some "researchers" are pleased with a big collection of names. If someone wants to use Rootsweb as a source that's fine. One still has to be mindful that there is a difference between a reference in Rootsweb to the California Death Records and actually reviewing the relevant record yourself. Even official records can be wrong. When an older person dies the people reporting the death may not have accurate information about the decedent's origins. I'm sure that many birth records include questionable paternity. I've had occasion to review the efforts of a recognized "professional" genealogist that got the generations mixed up. The report had been prepared to support a claim by an individual to status as a Native American. The information was used by a federal museum in Laramie, WY to misidentify a person in a photograph. This is much deeper than just a Wikipedia problem. Erroneous reportings of information can endure for centuries. When The Borg assimilates a whole civilization it also assimilates its errors. Can you personally afford to be bothered by this? I'm sure that the OED will find ways of dealing with words which have joined the vocabulary I certainly had the impression from chatting with Erin McKean at Wikimania that she is fully aware of the issues. Her "More Weird and Wonderful Words" (published by OUP) includes a "Webliography" as well as a more usual bibliography. The practice of adding quotes as evidence did not begin with the OED. My 1847 Webster and my 1817 Johnson have them. I don't think that any strict general rule about either using or not using web sources would be meaningful. This will be very disappointing to those who can't live without clear answers, and who will latch on to specific answers as a reason for excluding the opposing POV. Ec

On 1/4/07, Steve Block &lt;steve.block(a)myrealbox.com&gt; wrote: Professionally published dictionaries use those kind of sites as primary sources (examples of contemporary usages) in academic research about the usage of the English language. In other words, exactly the sort of thing we [[WP:OR|don't do]]. -- Robth (http://en.wikipedia.org/wiki/User:Robth)

Stan Shebs wrote: Funny, I thought they used it to document the English language. Maybe documenting the language is academic research though. -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.432 / Virus Database: 268.16.5/616 - Release Date: 04/01/07 13:34

On 03/01/07, Angela &lt;beesley(a)gmail.com&gt; wrote: I quite agree. Once an article is deleted it is no longer available to anyone who doesn't have admin status. There doesn't seem to be a co-ordinated effort among admins to sift through deleted articles to find useful information and so this information is effectively lost. Wikipedia's primary goal is to create free encyclopaedic content. IMO, it is always more useful for us to rewrite an article than to delete it wherever possible. If it is the case that the content is not suitable for Wikipedia (which, as long as it has a certain level of verifiability, is rare) then we should aim to put the content somewhere else so that it remains usable and the free content movement as-a-whole is benefited. -- Oldak Quill (oldakquill(a)gmail.com)

On 03/01/07, Bogdan Giusca &lt;liste(a)dapyx.com&gt; wrote: I don't really understand why an article would be useful for Internet Wikia but not useful for us... Their aims seem to be encompassed by our aims entirely. I go to Internet Wikia to find some information out about a website (eg. http://www.internet.com/). I want factually accurate, verifiable information. Why shouldn't I use Wikipedia in this case? I know that information from Wikipedia will be factually accurate and verifiable. I know that Wikipedia does not operate within the constraints of a traditional encyclopaedia so I hope to find at least a little information about that website on Wikipedia. The very purpose of Internet Wikia (providing "encyclopaedic" articles about presumably major websites) could, and should, be served by Wikipedia. Information is always enriched by being part of Wikipedia because it links to information and articles on unrelated fields. Conversely, Wikipedia is always enriched by having more information (within certain constraints). -- Oldak Quill (oldakquill(a)gmail.com)

Guy Chapman aka JzG wrote: While we can cut and paste a transwikied history to another project, that doesn't help if access to the diffs no longer works. It may satisfy the technical requirements of GFDL, but much information that might be useful in a new article is lost. Ec

Wednesday, January 3, 2007, 8:56:12 PM, Ray wrote: Too few on "good" subjects. And those that are on good subjects, are not properly written. Also, probably up to a quarter are created by their owners. Wikipedia still requires verifiability and reliable sources, right?

Wednesday, January 3, 2007, 10:38:55 PM, Jeff wrote: Imagine the spam we'd get then as "references". You know that a link from Wikipedia is pretty valuable for one's Google PageRank.

This is my opinion about references: Each one to its own. What that means is that depending on the topic, we should use sources that are appropriate. Something like science and history should typically always use peer reviewed publications, however webcomics do not have peer reviewed publications about them. That would be an appropriate time to invoke sources that would be more unorthodox for traditional scholarly subjects like science: blogs. I, for one, would trust a web comic blog about web comics because that's the best there is. Likewise, just because something is appropriate for one subject, doesn't mean it's appropriate for the next. If a web comic blog says that George Washington crossed the Delaware and defeated the Hessians, we shouldn't cite that web comic in our article about George Washington. This brings up another point of debate: using primary sources. There are times when primary sources are appropriate, and there are times when it's not. When quoting what someone famous said, the primary source is *the best* source for that information. That's what primary sources are for: pure facts (like what someone said). And then there are the times when secondary sources are more appropriate. Use those, then. There really isn't a source, or type of source, that is perfect for every kind of knowledge imaginable. Each one to its own. On 1/3/07, Bogdan Giusca &lt;liste(a)dapyx.com&gt; wrote:

On Wed, 3 Jan 2007 14:38:55 -0600 (CST), "Jeff Raymond" &lt;jeff.raymond(a)internationalhouseofbacon.com&gt; wrote: Yes, you'll get rid of all those tiresome people who think Wikipedia is an encyclopaedia and aspire to some kind of credibility, and you can get on with the job of documenting crap off teh internets, a genre for which there is a real paucity of verbiage based on other crap off teh internets. Wait, I think you just invented Fark. Guy (JzG) -- http://www.chapmancentral.co.uk http://en.wikipedia.org/wiki/User:JzG

On 1/3/07, Guy Chapman aka JzG &lt;guy.chapman(a)spamcop.net&gt; wrote: On good days, this keeps nutsy cruft away. On bad days, people wave the knife around to the extent that even The New York Times and books published by reputable publishers aren't sufficiently reliable, and the unpopular topic of the day gets nuked anyways. Consistency apparently isn't our strong suit so far this year... -- -george william herbert george.herbert(a)gmail.com