Could a subtractive analysis be used to test NPOV?
While WikiScanner is an important tool that advances critical inquiry regarding Wikipedia, it is not exactly user friendly. In addition, by acting as a catalog of edits with little information attached in its results list, the tool falls short in its most important task: keeping the public well-informed of what is being changed. In my opinion, WikiScanner currently focuses too much on the who and when of these changes
What if WikiScanner (or it’s data) could be adapted to display versions of Wikipedia pages that simply do not contain any edits originating from specific sources? What would a CIA-free Wikipedia look like? Would it expose tacitly accepted biases in the Wikipedia editors’ execution of “NPOV”? Could an instance of extreme yet easy to view whitewashing from a politician’s own staff ever be enough to sink their whole campaign?
The technique could prove tricky to implement. It would be imperative to mark where “censored” edits actually exist in the text, and to easily allow access to those edits in order to critically judge the merits and attributes of them as contributions. Even small changes in word choice can have a large effect on a so-called NPOV, so being able to view “censored” edits in relation to other edits (previous or attempted) would also be an important feature.
By providing tools to intuitively examine the contrast of a Wikipedia both with and without specific contributors, perhaps disingenuous parties would no longer find it politically prudent to attempt to hijack Wikipedia for their own ends.