Two arguments are frequently made which I think have no merit.
1) "It interferes with readability of the articles." Sure it does, but that's a technical issue, and _if we wanted solutions_ we could deal with it through tehnical means. We could define better visual apparatus for references. One of the examples I keep pointing to is Laura Hillenbrand's book, "Seabiscuit: An American Legend," not because it's the only book that uses this style but because it is an extremely readable bestseller. It is densely cited, but there is not a single mark within the text. Instead, the references are placed at the end. They are indicated by chapter, page number, _and phrase_.
This would require modifications to work with Wikipedia, but that can be discussed.
Wikipedia has a unique _requirement_ for very dense references, _denser_ than those found in research papers or nonfiction books, so it is not surprising that traditional solutions are not perfect for Wikipedia, and that we will need to think of better approaches.
2) "If it appears in numerous textbooks it does not need a citation." This is silly. The problem is that there is no way the reader or anybody else can tell the difference between a sentence which lacks a reference _because somebody has checked_ to make sure that it appears in numerous textbooks, and a sentence which lacks a reference because_ someone just typed it in off the top of their head_. They look the same.
Even if someone goes over an article with a fine-toothed comb today and has made sure that none of the unreferenced material needs references, without any sort of markup apparatus there's no way anyone can tell a week later which portions of the text have been reviewed.
In other words, if we don't drop something into the article to leave a breadcrumb trail to where the fact was found, then any work we do in fact-checking will be wasted effort because it will be obsolete a week later. We need some kind of marker to where the fact was found. And the marker needs to be readily visible... at least to those interested in seeing it... so that anyone can see how carefully the article has been fact-checked and _which facts in it_ have been checked.
Finally, which is easier to do: check to make sure that a fact is contained in _three_ textbooks and say "good, it doesn't need a reference" and not put one in? or check to make sure it's in ''one'' textbook and cite the source?
I sometimes think that at least some people who object to citations do so because what they really want is to _establish themselves as authorities_ through social interaction with other page editors. That is, they want their Wikipedian colleagues to recognize _them_ as reliable sources, and agree that any fact inserted by [[User:Pantomath]] does not need a citation because everyone agrees that User:Pantomath knows everything.
On 9/29/06, Daniel P. B. Smith wikipedia2006@dpbsmith.com wrote:
Wikipedia has a unique _requirement_ for very dense references, _denser_ than those found in research papers or nonfiction books, so it is not surprising that traditional solutions are not perfect for Wikipedia, and that we will need to think of better approaches.
- "If it appears in numerous textbooks it does not need a citation."
This is silly. The problem is that there is no way the reader or anybody else can tell the difference between a sentence which lacks a reference _because somebody has checked_ to make sure that it appears in numerous textbooks, and a sentence which lacks a reference because_ someone just typed it in off the top of their head_. They look the same.
Yes, yes, yes. Our articles need to come in with a chip on their shoulders, as it were. There is no external reason to believe that what it says resembles the truth, no author or organization claiming responsibility for the text. The article needs to make a case for the information it presents being accurate. Don't just state facts, show the reader how they can confirm the accuracy of the statements. Trusting a Wikipedia article requres a leap of faith; we want to minimize the distance of that leap.
Daniel P. B. Smith wrote:
- "If it appears in numerous textbooks it does not need a citation."
This is silly. The problem is that there is no way the reader or anybody else can tell the difference between a sentence which lacks a reference _because somebody has checked_ to make sure that it appears in numerous textbooks, and a sentence which lacks a reference because_ someone just typed it in off the top of their head_. They look the same.
They certainly don't look the same, unless the person "reviewing" the article utterly lacks any competence to review the article, in which case they should kindly refrain from doing so. Anyone who has even very basic competence knows what is an uncontroversial statement that appears in numerous textbooks in their field.
I don't agree with the expert-centric approach Larry Sangers advocates, but we don't have to have only people who have *no* idea what they're doing editing our articles either.
-Mark
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Delirium wrote:
They certainly don't look the same, unless the person "reviewing" the article utterly lacks any competence to review the article, in which case they should kindly refrain from doing so. Anyone who has even very basic competence knows what is an uncontroversial statement that appears in numerous textbooks in their field.
I don't agree with the expert-centric approach Larry Sangers advocates, but we don't have to have only people who have *no* idea what they're doing editing our articles either.
-Mark
We are talking about /readers/ here, not Wikipedia editors. If a /reader/ sees a particular unreferenced statement on Wikipedia, they do not know whether the statement is unreferenced because: a) Some idiot has made it up without any factual basis b) It is 'widely accepted as a fact'
Your 'people familiar with the field will recognise it as a well-known fact' argument doesn't really wash in this situation. The reason a person would be reading a Wikipedia article on a subject is because they _don't_ know enough about the subject. What may appear as a 'widely accepted fact' to someone writing a Wikipedia article on a subject may be nothing of the sort to an uninformed person seeking to use Wikipedia to expand his/her knowledge (which is, after all, what we're for).
Cynical
David Russell wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Delirium wrote:
They certainly don't look the same, unless the person "reviewing" the article utterly lacks any competence to review the article, in which case they should kindly refrain from doing so. Anyone who has even very basic competence knows what is an uncontroversial statement that appears in numerous textbooks in their field.
I don't agree with the expert-centric approach Larry Sangers advocates, but we don't have to have only people who have *no* idea what they're doing editing our articles either.
-Mark
We are talking about /readers/ here, not Wikipedia editors. If a /reader/ sees a particular unreferenced statement on Wikipedia, they do not know whether the statement is unreferenced because: a) Some idiot has made it up without any factual basis b) It is 'widely accepted as a fact'
Your 'people familiar with the field will recognise it as a well-known fact' argument doesn't really wash in this situation. The reason a person would be reading a Wikipedia article on a subject is because they _don't_ know enough about the subject. What may appear as a 'widely accepted fact' to someone writing a Wikipedia article on a subject may be nothing of the sort to an uninformed person seeking to use Wikipedia to expand his/her knowledge (which is, after all, what we're for).
Well, that's why I think we *should* have review processes of some sort. Then if a reader sees an unreviewed article, they are indeed in the situation you describe. But if a reader sees an article that has been reviewed as at least "pretty good" by people who know something about the field, then they can rest assured that nothing obvious is wrong about it. For what it's worth, I do think we should have a different process of reviewing than good/featured, but that's been talked about at length before.
In any case, given references doesn't solve the problem you describe either, because then the reader still can't determine without some background knowledge whether this is a mainstream/reliable reference, or a reference to a book written by someone with a decidedly minority view. That's why I think we should help readers out by having people with at least some knowledge in the area the article's about give the reader some indication of how good the article is---delegating all the fact-checking to every individual reader simply doesn't work.
-Mark
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Delirium wrote:
Well, that's why I think we *should* have review processes of some sort. Then if a reader sees an unreviewed article, they are indeed in the situation you describe. But if a reader sees an article that has been reviewed as at least "pretty good" by people who know something about the field, then they can rest assured that nothing obvious is wrong about it. For what it's worth, I do think we should have a different process of reviewing than good/featured, but that's been talked about at length before.
(snip) -Mark
Yeah but people still wouldn't know if the article had changed since it was deemed 'pretty good'.
Cynical
Delirium <delirium@...> writes:
David Russell wrote:
While I support the option of expert review of Wikipedia article revisions, I think Delirium misunderstood the point of referencing. ...continued below...
Delirium wrote:
They certainly don't look the same, unless the person "reviewing" the article utterly lacks any competence to review the article, in which case they should kindly refrain from doing so. Anyone who has even very basic competence knows what is an uncontroversial statement that appears in numerous textbooks in their field.
I don't agree with the expert-centric approach Larry Sangers advocates, but we don't have to have only people who have *no* idea what they're doing editing our articles either.
We are talking about /readers/ here, not Wikipedia editors. If a /reader/ sees a particular unreferenced statement on Wikipedia, they do not know whether the statement is unreferenced because: a) Some idiot has made it up without any factual basis b) It is 'widely accepted as a fact'
Your 'people familiar with the field will recognise it as a well-known fact' argument doesn't really wash in this situation. The reason a person would be reading a Wikipedia article on a subject is because they _don't_ know enough about the subject. What may appear as a 'widely accepted fact' to someone writing a Wikipedia article on a subject may be nothing of the sort to an uninformed person seeking to use Wikipedia to expand his/her knowledge (which is, after all, what we're for).
Well, that's why I think we *should* have review processes of some sort. Then if a reader sees an unreviewed article, they are indeed in the situation you describe. But if a reader sees an article that has been reviewed as at least "pretty good" by people who know something about the field, then they can rest assured that nothing obvious is wrong about it. For what it's worth, I do think we should have a different process of reviewing than good/featured, but that's been talked about at length before.
In any case, given references doesn't solve the problem you describe either, because then the reader still can't determine without some background knowledge whether this is a mainstream/reliable reference, or a reference to a book written by someone with a decidedly minority view.
Not exactly, no. To determine if a fact is "well known to be true" requires subject specific knowledge; to determine if a source is mainstream requires more general skills at judging sources *in general*; while subject familiarity is one way to judge a source's reliability it is not the only method. With the exception of deductive or indutive logic, subject specific knowledge is the only way to verify a specific fact. That's why I think we should help readers out by having people
with at least some knowledge in the area the article's about give the reader some indication of how good the article is---delegating all the fact-checking to every individual reader simply doesn't work.
In any case, having expert review or not is a distraction from the purpose of referencing. The most important point of referencing, for our readers, is to allow them to tell *where* our claims come from. Even if a lay reader can't tell if "Prof. John McManins, published in the National Dynamic Society Journal" is a reliable source, if they tell someone who does know the subject - "X is true because McManis said it." the knowledgable person will have something to go on. If they just say "X is true because Wikipedia said it." this is useless; there's nothing further to say. Referencing encourages our readers to pay attention to where something comes from, not just take it on faith. This is critical.
References shouldn't be *required* - I think the GA and FA requirements of references are more or less misplaced (esspecially in the case of GA), but any amount of references should be *allowed*, and appreciated.
Jesse Weinstein
Hi Mark, All,
Delirium delirium@hackish.org wrote:
They certainly don't look the same, unless the person "reviewing" the article utterly lacks any competence to review the article, in which case they should kindly refrain from doing so. Anyone who has even very basic competence knows what is an uncontroversial statement that appears in numerous textbooks in their field.
Just stumbled over another example: http://en.wikipedia.org/wiki/Renaissance
There were also other arguments for not promoting to GA right now (e.g. relation to the [[Italian renaissance]]) article) but prominetly cricized was the lack of in-line cites.
Heck, this is an overview article. It should only contain current widespread consensus to be found in major textbooks.
Yes, as specific scholars are mentioned, their works can get a footnote, but that's not primarily a WP:V issue.
Regards, [[User:Pjacobi]]
On 30 Sep 2006, at 03:09, Daniel P. B. Smith wrote:
Wikipedia has a unique _requirement_ for very dense references, _denser_ than those found in research papers or nonfiction books, so it is not surprising that traditional solutions are not perfect for Wikipedia, and that we will need to think of better approaches.
We also have more solutions at our disposal.
May preference is to use colour. There are many possible ways to do this, but one way would be for sentences to start out light grey, and each independent editor who approves or disapproves turns it darker and either more blue or more red. So dark blue would be strongly supported, dark red would be strongly opposed, and light grey would be not supported or opposed.
Then uncited claims which were "obviously" true would soon gain credibility, but contentious claims would also be obvious.
As has been mentioned, a simple viewing option could turn everything back to normal.
On 9/30/06, Stephen Streater sbstreater@mac.com wrote:
May preference is to use colour. There are many possible ways to do this, but one way would be for sentences to start out light grey, and each independent editor who approves or disapproves turns it darker and either more blue or more red. So dark blue would be strongly supported, dark red would be strongly opposed, and light grey would be not supported or opposed.
Then uncited claims which were "obviously" true would soon gain credibility, but contentious claims would also be obvious.
As has been mentioned, a simple viewing option could turn everything back to normal.
Please, no!
This is the sort of "voting for the truth" that Colbert dubbed "wikiality".
It is not a question of how many people think it's true, or vote for it to be true. It's a question of reliable, verifiable sources.
Yes, we can go wonko about process, or smother articles with reference-love (neither of which is good), but let's not encourage the view that whatever most people believe must be the truth.
-Rich [[W:en:User:Rholton]]
On 30 Sep 2006, at 18:33, Richard Holton wrote:
On 9/30/06, Stephen Streater sbstreater@mac.com wrote:
May preference is to use colour. There are many possible ways to do this, but one way would be for sentences to start out light grey, and each independent editor who approves or disapproves turns it darker and either more blue or more red. So dark blue would be strongly supported, dark red would be strongly opposed, and light grey would be not supported or opposed.
Then uncited claims which were "obviously" true would soon gain credibility, but contentious claims would also be obvious.
As has been mentioned, a simple viewing option could turn everything back to normal.
Please, no!
This is the sort of "voting for the truth" that Colbert dubbed "wikiality".
It is not a question of how many people think it's true, or vote for it to be true. It's a question of reliable, verifiable sources.
Yes, we can go wonko about process, or smother articles with reference-love (neither of which is good), but let's not encourage the view that whatever most people believe must be the truth.
There is a difference between voting for the truth and showing which statements have been widely considered.
On 9/30/06, Stephen Streater sbstreater@mac.com wrote:
On 30 Sep 2006, at 18:33, Richard Holton wrote:
On 9/30/06, Stephen Streater sbstreater@mac.com wrote:
May preference is to use colour. There are many possible ways to do this, but one way would be for sentences to start out light grey, and each independent editor who approves or disapproves turns it darker and either more blue or more red. So dark blue would be strongly supported, dark red would be strongly opposed, and light grey would be not supported or opposed.
Then uncited claims which were "obviously" true would soon gain credibility, but contentious claims would also be obvious.
As has been mentioned, a simple viewing option could turn everything back to normal.
Please, no!
This is the sort of "voting for the truth" that Colbert dubbed "wikiality".
It is not a question of how many people think it's true, or vote for it to be true. It's a question of reliable, verifiable sources.
Yes, we can go wonko about process, or smother articles with reference-love (neither of which is good), but let's not encourage the view that whatever most people believe must be the truth.
There is a difference between voting for the truth and showing which statements have been widely considered.
In what way would such a color-based interface reflect that difference? If there was a widely-accepted but false statement in an article, how would it appear based on this interface? How would such an interface help (or hurt) in a hotly-contested article, like [[Intelligent design]] or [[George W. Bush]]?
I just don't see such an interface contributing anything to the issues it attempts to solve.
-Rich [[W:en:User:Rholton]]
On 30 Sep 2006, at 20:52, Richard Holton wrote:
On 9/30/06, Stephen Streater sbstreater@mac.com wrote:
On 30 Sep 2006, at 18:33, Richard Holton wrote:
On 9/30/06, Stephen Streater sbstreater@mac.com wrote:
May preference is to use colour. There are many possible ways to do this, but one way would be for sentences to start out light grey, and each independent editor who approves or disapproves turns it darker and either more blue or more red. So dark blue would be strongly supported, dark red would be strongly opposed, and light grey would be not supported or opposed.
Then uncited claims which were "obviously" true would soon gain credibility, but contentious claims would also be obvious.
As has been mentioned, a simple viewing option could turn everything back to normal.
Please, no!
This is the sort of "voting for the truth" that Colbert dubbed "wikiality".
It is not a question of how many people think it's true, or vote for it to be true. It's a question of reliable, verifiable sources.
Yes, we can go wonko about process, or smother articles with reference-love (neither of which is good), but let's not encourage the view that whatever most people believe must be the truth.
There is a difference between voting for the truth and showing which statements have been widely considered.
In what way would such a color-based interface reflect that difference? If there was a widely-accepted but false statement in an article, how would it appear based on this interface? How would such an interface help (or hurt) in a hotly-contested article, like [[Intelligent design]] or [[George W. Bush]]?
I just don't see such an interface contributing anything to the issues it attempts to solve.
Perhaps you could continue to look a little deeper.
A false but widely accepted item would show up as having consensus agreement. A discussion on the talk page by the enlightened few would be expected to change this consensus - possibly including a link to the discussion which included any cites so new people could see how the consensus was reached. This compares to the current system where such a statement sits there without comment.
A false but not widely supported view, on the other hand, would be much more distinct. In the current system, it appears black and indistinguishable to the lay reader from a widely known truth. In the colour system, such an unsupported opinion would appear as pale grey.
Even hotly contested articles contain many facts agreed by all sides. The contested information and the agreed information would be clearly distinguished. This is an area where many cites to support any opinion would be available, so citations on their own would not resolve the differences in the current set up.
So the issues it would resolve are (a) to clarify which areas have few or many opinions (b) to clarify which areas are contentious