User:Pgallert/OR–22

OR–22 is the catch-22 situation arising from a group of related Wikipedia policies with the policy on Original Research (OR) at the core of the problem. This essay claims that it is impossible to adhere to a strict reading of OR, shows examples of where OR essentially must be violated, and offers a way out of the dilemma by suggesting a different core wording.

One of the Wikipedia policies most frequently violated by experienced editors is No original research (NOR). As a result, Wikipedia is full of unattributable claims. This, however, is unavoidable because it is impossible to fulfill the policy on Original Research and the guideline on Plagiarism at the same time.

Problem statement
I have a reliable, independent source that I want to use to expand an article on Wikipedia with. The text in the source is under copyright. Now I have the following problem:
 * 1) If I use the text as-is, I commit plagiarism.
 * 2) If I change the wording ever so slightly so that the text keeps its intended meaning, I commit plagiarism.
 * 3) If I change the wording so much that the meaning of the text changes, I violate the Original Research policy, among others.

I can only do 1, 2, or 3. Any substantial rewording of a text will inevitably change its meaning slightly. This might be intended or not, and might go unnoticed or not, but it is unavoidable.

Experienced WP editors tend to choose option 3.

Examples of unattributable claims
I have collected a few examples to illustrate my point. All of them are from my own editing history because I do not want to implicate other editors. I do not believe that I violated common sense in the examples below.

Pictures and picture subscripts
The subscript under pretty much any picture taken by a non-subject expert is original research. Take the school of Seeis: You cannot say if this group of buildings is a school, nor if it is situated at Seeis or anywhere else in the savannah. I took the picture and can vouch that it indeed is situated at the place it is claimed to be. This may not be sufficient as I am not an expert on geography. But if it is indeed a school even I cannot say for sure; a villager told me so, and I took it for granted. Needless to say that it is extremely unlikely that someday a reliable source is produced that shows a similar picture and likewise claims it is the school of Seeis. The fact is essentially unattributable, and that it will likely never be challenged is only due to the situation that nobody cares about Seeis.

Cherry-picking of sources
A selective choice of references to back up article claims is among the most powerful weapons of POV-pushers. But even in neutral territory it is part of the daily business of en editor. Moreover, it is a "crime" that is hard to prove, after all I could have overlooked certain sources and by chance only found the ones I have used for the article.

Sometimes it is necessary to exclude a reliable, independent source on purpose. This is when we get into OR territory. Omitting statements is an action, done consciously and with full intent.

This weekly column is normally a rather reliable source. It appears in a reasonably independent national newspaper, it covers one local hero at a time, and I have written many Wikipedia articles based on this article series.

However, for Munjuku Nguvauva II I chose to ignore this source when writing that Nguvauva did not attend the Turnhalle Constitutional Conference. It was clear (to me---OR!) that the author based his article entirely on the work of Klaus Dierks, and it was clear (to me---OR!) that the author misread Dierks' statement about the Turnhalle Conference. It is important to note that this disagreement is not one that can, or should, be documented in the article in the form A states that..., whereas B opines that.... It is only one scientist and one journalist disagreeing, and they do not form opposite laagers: The journalist simply misunderstood what the scientist had said.

Another case is that of missionary and historian Heinrich Vedder. He has written a number of books, the most important one (OR!) is Over the years I have encountered many inconsistencies in this book, and I have heard claims (OR!) from colleagues that Vedder was more of a story-teller than a historian. I have no reputable, published reference for this suspicion. Still, whenever Vedder's version disagrees with some other historic account, I give preference to Vedders opponents, for instance in the history sections of Gobabis and Leonardville. This is actually absurd. Vedder wrote the only book on early Namibian history. Different views come from the statements of farmers, missionaries, and tribal chiefs---people that are normally a lot less trustworthy than historians. But if Vedder's narrative makes no sense, and another one does, I think I provide a service to Wikipedia to ignore the former.

Implicit evaluation
Sometimes the way a source is represented in an article passes a judgment on the value of that source. When writing Reiterdenkmal, Windhoek I was in the situation that a respected historian makes an unusual claim: The monument is unique in that it displays a corporal on horseback. Nowhere else in the world is an ordinary soldier sculpted in that manner; the honour to be displayed on a horse is only extended to nobles and rulers.

I have a reference that supports this claim, but I thought it to be unlikely that a local historian has the time and the means to check all other equestrian statues to have his claim verified. I also took into consideration the heated atmosphere in which Vogt's claim was published. It was weeks before this statue was to be removed and very likely damaged in the process. So I added Historian Andreas Vogt claims that nowhere in the world... While this assertion is factual, the wording now suggests that Vogt might be wrong with his claim. This is OR, I have no reference supporting this suspicion. And it is a strong claim, because it implicitly states that a historian who is an expert on local Namibian monuments makes a wrong claim in his area of expertise. In its current version, the phrase even reads: Historian Andreas Vogt incorrectly claims...

Again, without a reference. The word incorrectly was inserted after a number of editors claimed that other monuments likewise depict foot soldiers.