Editorialized Language Imperils Trust in Fact-Checking
When it comes to improving fact-checking, most of the focus tends to be on how fact checkers select the claims they examine, the sources they turn to in determining “truth” and the strength of their arguments. While it receives less attention, the editorializing language fact checkers use can be just as harmful to the public’s trust in their work.
Browse through the first five pages of FactCheck.org’s Barack Obama archive and fact checks of the former president’s statements are headlined with relatively clinical terms such as “misrepresents,” “misleading” and “cherry-picks.”
In contrast, the first five pages of the site’s Donald Trump archive are headlined “bungles,” “whoppers,” “deceptive,” “tramples facts,” “distorts,” “wrong again,” “false,” “wrong” and “errors,” with past headlines using “bizarre,” “baffling” and “fanciful.”
What separates a “misrepresented” fact from a “bungled” one or a “misleading” statement from a “whopper”?
Asked how FactCheck.org determines the wording it uses in its headlines, Eugene Kiely, the site’s director, offered that “we write headlines that we think accurately reflect the story” and that they “are based on the evidence” and “reflect the facts” of each individual story.
Does this mean a “whopper” reflects a more serious falsehood than a “cherry-picked” one? Kiely clarified that “we don’t write headlines based on the severity of the error” and that the language used “isn’t intended as a rating system.”
As to how the site selects words to use if they are not based on severity, Kiely said only, “We do NOT have a rating system. We do NOT choose words as a way to rate the severity of the misinformation, but rather to most accurately describe it.”
Asked for a copy of its written guidelines governing these matters, he referred back to his earlier statements that headline wording is selected on a case-by-case basis to fit each story.
This lack of a formal policy outlining the criteria under which “bizarre” should be used instead of “misspoke” raises questions about why the site believes such charged language would not spark neutrality concerns.
In the absence of objective guidelines, a reader confronted with clinical language describing Obama’s falsehoods alongside sensationalized and inflammatory descriptions of Trump’s might reasonably question whether the fact checker has a political bias.
Asked about these concerns and why FactCheck.org does not stick with clinical language for all its headlines, Kiely responded, “I have no concern about using language that accurately reflects the evidence.”
In the end, the use of catchy clickbait headlines risks eroding public trust in the fact-checking enterprise by creating the impression of bias. In the place of “whoppers” and “bungled” facts, sticking with “incorrect” and “misleading” would go a long way toward eliminating any suspicions of it.