|This page is currently inactive and is retained for historical reference. Either the page is no longer relevant or consensus on its purpose has become unclear. To revive discussion, seek broader input via a forum such as the village pump.|
Wikipedia should be consistent. There are two kinds of consistency:
- factual – statements in articles shouldn't contradict each other,
- stylistic – the organization and presentation of the information should be uniform across articles.
This proposal deals with the factual consistency only. The stylistic consistency is thoroughly specified at Wikipedia:Manual of Style.
It is inevitable that same facts are repeated in more than one place in Wikipedia, generally for two reasons:
- Every article needs to establish context and to be easy to read. This is impossible without repeating information that exists in other articles.
- There are additional navigational articles, such as disambiguation pages, lists, glossaries and comparisons.
It seems that every factual statement on Wikipedia can be classified as:
- primary – statement about the subject of the article, and if this information would be included in Wikipedia exactly once, there would be no better place to put it than this,
- secondary – statement about object of another article, such as comparison or context.
All the factual information in navigational articles should be considered as secondary statements. This is similar to distinction between primary and secondary sources, but restricted only to Wikipedia.
In order to have factual consistency, the following policy is proposed:
Primary factual statements should occur on Wikipedia only once, and every other repetition should be the secondary factual statement. The primary statements are authoritative, and the changes (even correct and/or sourced) to secondary factual statements should be accepted after the primary statement in question is changed.
The rationale behind this is that there is probably more editors knowledgeable about the topic around the page that contains the primary statement. So to ease the verification, the primary stamenents should be verified, but the secondary statements can refer to already verified (or assumed to be verified) primary statements.
Practical example: Recently, on page Comparison of programming languages, someone noted that Java is has been influenced by C#. I (User:Samohyl_Jan) reverted the edit, because it contradicted what Java page itself said about the subject, even though the statement was sourced (and I partially agreed—it depends on the definition of influence), because this should be correctly on Java page first and the comparison page should refer for the factual statement. Then, after this has been accepted on Java page, it is only logical to have it the same on the comparison page.
- This was wrong. Referenced statements should not be removed. What you should have done was update the Java page instead of removing the sourced statement. Joelito (talk) 19:59, 26 November 2006 (UTC)
- Perhaps you're right. What I was afraid of is this: the language comparison page is pretty obscure page with respect to Java language page. If this statement were wrong (I wasn't quite sure) and I would put this to Java page too, it could have been reverted, but then it could stay on language comparison page. I don't mind if the correct procedure should be to propagate the new change so the more people could tell if it is right or wrong, but in my opinion (as I explain above by division to primary and secondary facts on Wikipedia) this could be more wrong than to revert the change in obscure place. Samohyl Jan 22:50, 26 November 2006 (UTC)
I believe such a rule can make reverting of vandalism and resolving disputes a lot easier, even though it can sometimes lead to more errors.