A Cautionary Note

Fact-checking is an enterprise that is bound to attract challenge, from scholars as well as political actors. LiarLiarLiar does not itself serve as a fact-checker — our role is instead to make it easier for you to choose from and use a broad range of existing fact-checkers. But transparency is a core value for us. So we must address the doubts people have sometimes had about fact-checking and related services in general.

We believe it is useful to distinguish between three categories of doubt. First, at least some of the doubts that are commonly raised about fact-checkers are easily dismissed, as products of “interested parties” who have been caught out by fact-checkers and hope to undermine the source of their embarrassment. These statements are themselves subject to fact-checking, however. So they can be tallied by the same mechanisms we offer for other purposes, and weighed against other statements.

A second category of doubt about fact-checkers themselves is more general: almost everyone is inclined to trust some organizations more than others, and some may have very strong preferences in this area. LiarLiarLiar was designed from the beginning with this in mind: You, not we, decide which organizations’ fact-checking to trust, for a given search or all your searches. And it’s not a binary trust/distrust choice; you also get to decide how much you want to rely on one source relative to others. We undertake to list not just the most commonly known fact-checkers but also others that meet a basic standard of integrity, as defined in turn not by us, but by nonpartisan groups set up for that specific purpose. And we name those organizations, so you will know the chain of logic and trust we ourselves are following.

There is, however, a third and more subtle category of doubt. A distinguished sociologist of our acquaintance put it this way: “Fact-checking tends to take the questions for granted rather than reflect on how the question comes to be asked that way. What answer is implicit in the framing of the question?”

Our colleague also notes two related problems: that “Fact checkers necessarily believe that there are facts independent of human purposes, which reflects a particular theory of ontology. [LiarLiarLiar] limits the universe of comparisons to fact checkers recognized by other fact checkers through their membership of fact-checking associations”, and that “the problem that [any such engine] ultimately confronts is that of deciding when a fact is a fact rather than an opinion backed by an exercise of power.”

These are real concerns, and we are indebted to Robert Dingwall for his thoughtful help in highlighting them. We believe the best response we can make is to be pragmatic: Rather than abandon offering any such service, we must admit the results will never be perfect. Any fact-checking mechanism is expensive to set up, and must be operated by some kind of organization. Even halfway reliable sources are therefore limited in numbers and scope, and perhaps point of view as well. One more reassuring factor is that much of the misinformation and disinformation that circulates so widely appears not particularly subject to distinctions as to how exactly a question may be framed, but instead relies on quite blatant lying.

There is certainly room for other approaches, and we hope they will be developed in future. There is also plenty of room for a broader array of organizations to set up their own fact-checkers; perhaps the easier access offered by a service such as ours will help encourage more groups to undertake that work. When new fact-checkers are offered, we undertake not to reject them out of hand, even if they do not join the existing fact-checker associations we are currently relying on.

But for our purposes and for the time being, we believe democracies are better off with tools that are admittedly imperfect than without them. For the rest, we can only emphasize that we make no claim that LiarLiarLiar is a one-stop-shop for some ultimate definition of Truth. 

Find Out More

(to link to a more thorough account, or even to a proper scholarly treatment of epistemology.)