3 ms·
The report quality is excellent and a worthwhile template for disinformation investigations in other contexts. I found it especially interesting because while I
by anigbrowl 5y ago
The report quality is excellent and a worthwhile template for disinformation investigations in other contexts. I found it especially interesting because while I am not an expert in Balkan affairs, I'm expert-adjacent; have friendships with refugees, journalists and former military/intelligence involved in the conflict on various different sides and have read deeply into it in my own time.
A standout was the table listing relevant facts (specifically convictions of war criminals by an international tribunal) and whether the language sub-wikis accurately reflect those facts or not. Whether one likes or dislikes a given fact, its occurrence is a kind of ground truth that can serve as a reference marker by which to weight the factual content of an article. Thus, whether someone was convicted by a tribunal is a fact; whether it was fair is open to debate, but failure to mention the fact at all amounts to a concealment of information, whether negligent or willful. When we see a pattern of factual errors (including glaring omissions), we can assess the degree of deliberation involved by looking at whether the errors all run in similar directions and how systematic/assiduous the erroneous editor is in rejecting alternative points of view.
A complicating factor here is Wikipedia's general open-access policy. As you probably know, some people who are interested in a topic appoint themselves gatekeepers of it and revert edits they disapprove of. Over time, I can be difficult the distinguish between passion and expertise. For example, suppose you grew up in a village where a battle was fought (before you were born or aware of such things). You receive from your parents, peers, and teachers an approximate view of the historical event as experienced by those who lived through or were directly impacted by it. While in high school, you find a Wikipedia page about the battle which has some details wrong - geographical facts, say - and you correct those with your local knowledge. As time goes on, you become the leading Wikipedian curating this topic, and enrich the article with all kinds of valuable information on how the battle was fought.
But suppose you find yourself getting into arguments about why the battle was fought; now you're going outside the scope of your local knowledge. Battles might fought for purely logistic reasons (control a river crossing, cut an enemy off from supply lines), personal ones (a commander has family or reasons to love or hate a particular place), or strategic factors (to weaken enemy resolve or tempt a misallocation of forces). These might not be apparent from the local perspective; received wisdom might posit a local cause that was actually just the initial skirmish of a larger action.
Case studies illustrate some of the difficulties in evaluating controversial information. For example, the entry on Ratko Mladić, a Bosnian Serb, is heavily slanted on Serbian wikipedia, which emphasizes his positive qualities and plays down negative aspects of his story, as well as introducing analysis and speculation favorable to Serbian nationalist narratives. Now, a problem with bias accusations is subjectivity; a Mladić apologist might argue that other narratives omit contextual information that seems to justify his actions, or that emphasis on his criminality serves to distract from crimes by other parties that equally severe but less strenuously prosecuted.
Although there may be large differences of opinion, that can be OK when the contextual variations are truthful. Different people can have very different ideas about what the significant factors are, and exploring those differences can be valuable. Most historical arguments (in my non-professional view) are not so much debates about what happened but rather over what matters, and that can vary both because people are on different sides of a tribal argument but also because people have different assumptions about moral/ethical considerations that are orthogonal to their tribal membership.
Missing from the report (perhaps because most people are not that interested) is any kind of dynamic information about the development and intensity of the divergences. It might be interesting to mine the article history and/or talk pages to get a sense of which parties struggled over control and how those struggles played out - for example, was there an openly revisionist challenge on core topics that was eventually sustained by attrition of disagreeing voices, or a stealth campaign of incremental nudges in one direction with a large cumulative effect?Modeling and visualizing these dynamics could yield valuable insights on other sorts of networked conflicts.