Jump to content

Wikipedia:Trust, but verify

From Wikipedia, the free encyclopedia

There is long-standing consensus that Wikipedia is not a newspaper (especially not a tabloid), but this is in tension with the desire of Wikipedians, acting in good faith, to cover breaking news and developing events.

On the one hand, we are supposed to use secondary, i.e. analytical, sources. On the other, there can be breaking news where Wikipedia could play an important role in presenting the facts in an unbiased way, against a background of spin and speculation – even if the facts are "this is all speculation and nobody really knows".

This is particularly true of narratives driven out of ideological media bubbles. As Swift put it: "Falsehood flies, and the Truth comes limping after it". Steve Bannon described his tactics as "flood the zone with shit",[1] and more measured sources have noted that, for example, a hack-and-leak operation "is designed to spread far and wide long before all the real facts can be known".[2] It's notable that the 2017 Macron e-mail leaks had little effect on the election they were designed to throw, due to the French media's mandated 44-hour media blackout for elections.

Wikipedia's mission is verifiability, not truth, but that does not absolve us of responsibility for factual accuracy when it is clear that what is verifiably being said by a number of people is not verifiably based on any credible factual footing. Wikipedia should not breathlessly follow the lie, but should instead wait to see if the truth has donned its boots yet.[3] In short, we should trust but verify.

A good way to do this is to compare the factual (as opposed to comment) reporting in major news outlets that have different political leanings. If something is stated as fact by both the Wall Street Journal and the Washington Post, it is almost certainly safe to state it as fact on Wikipedia. If, on the other hand, there is a serious discrepancy between comment and factual sources, or if comment sources are the only source of a particular claim, then there is a very high probability that you're not hearing the whole story yet, and we probably should not touch it, especially if it concerns living people. And if the comment side and the factual side of a major news site disagree on the fundamentals underlying the claim, that's a huge red flag.

News sources are under threat from new media. The transfer of advertising revenue to online and new media has crippled the budgets of most news organisations. 98.5% of Facebook's revenue is advertising, a total of nearly $70bn in 2019. That pushes engagement-driven opinion high up the charts. Facebook and Alphabet (Google) now absorb the lion's share of advertising revenue in the US and many other countries. Ad revenue follows clicks, and clicks follow shares. The Twitter account @FacebooksTop10 tracks the top-performing links on Facebook and almost all are opinion. So news sources have thinned the line between news and opinion desks, which in turn reinforces bias. This is both universal and asymmetric. Yochai Benkler has used network analysis to track cross-references between media platforms, and came to the striking finding that over time the conservative media bubble has become more insular, more self-referential, and markedly less likely to reference mainstream sources, while the mainstream has continued to reference across a spectrum of ideology.[4]

AP, a source generally considered accurate and unbiased, started using the term "accountability journalism" to describe an increased tolerance for journalists telling the truth "as they see it" rather than striving for objective fact.[5] The AP's approach was readily adapted by many other mainstream reliable sources, aimed "to scrutinize and hold powerful people and institutions accountable" with increased fact-checking and investigative stories.[6] However, as both non-traditional means of news delivery on the Internet sprang up in the early 2000s, along with the rise of misinformation and fake news sources that took advantage of mimicking the accountability journalistic style, these sources have further verved into inserted more opinion and assert "moral clarity" into their reporting to try to keep the trust of their readership and stay commercially competitive and relevant in the current environment.[7][8] These broad moves have been criticized as abandoning the principles of objectivity that the press was generally recognized for in the past, and muddies how Wikipedia must use these sources; these sources are not suddenly unreliable, but we must be careful of taking each printed word in a news piece as uncontested fact.

We are generally good at distinguishing between good and bad sources for medicine and pseudomedicine. Despite occasional attempts at false balance we generally reflect peer-reviewed scientific findings as true and the claims of charlatans as false, and don't have a lot of difficulty distinguishing between them. That's not so true for current events, where the line between objective fact and ideological Truth™ can be difficult, if not impossible, to establish. Bias doesn't necessarily undermine the accuracy of the reported content, but it requires us to be more careful to look for language that frames a report as fact or opinion, and it suggests that at the very least we should be looking at sources with a variety of editorial positions and not relying on one source as the arbiter of fact about a current event. And if in doubt we must wait for more academic/long-term review coverage to provide the more detailed commentary.

For Wikipedia, the best solution is always to hold back until analytical sources are available. This will normally happen in good time for the deadline. Judgment of sources should be based on independent reviews of factual accuracy, we should not include a source because it presents one side or other of a dispute – especially for claims that have far-reaching effects. Sources must be reliable AND independent AND secondary, not any combination of the three.

Per WP:NOR we are not qualified to be arbiters of fact, only of significance. We cannot take two competing and mutually contradictory opinion pieces and discern the facts from them. We should not even try. If opinion is all there is, it's probably WP:UNDUE.

Recommendations

[edit]
  • Don't base articles on commentary sources.
  • Don't try to beat the chyron: wait for analytical sources.
  • Where politics are concerned especially, compare the factual reporting of left- and right-leaning sources.
  • The opposite of mainstream is neither conservative nor liberal, but fringe. Anything not covered in the mainstream media should probably not be covered on Wikipedia either, because we are a mainstream project.
  • Remember that neutrality is not the average between competing narratives. The reality of climate change is not the mid point between the IPCC and the Heritage Foundation.
  • Remember that in politics almost every source has a dog in the fight. And sometimes they are the dog.
  • When something is propagating rapidly but looks suspicious or implausible, the responsible thing is to exclude it from Wikipedia until responsible sources have analysed it.

References

[edit]
  1. ^ Time
  2. ^ Hack and leak v social media, Financial Times
  3. ^ You're going to say Mark Twain, but you're almost certainly wrong.
  4. ^ Benkler, Yochai (2018). Network propaganda : manipulation, disinformation, and radicalization in American politics. New York, NY. ISBN 978-0-19-092362-4. OCLC 1045162158.{{cite book}}: CS1 maint: location missing publisher (link)
  5. ^ Digital Journal
  6. ^ [1]
  7. ^ [2]
  8. ^ [3]