Truth: It’s Not What People Say It Is

Meta recently announced that it would end its use of independent fact-checkers, initially decommissioning the system in the United States, and, perhaps after that, elsewhere across the world.¹ Instead, the company will move to a "community notes" system, like that favoured by other social media platforms. This move produced a good deal of media debate,^2,3 some of which implies this is a bad thing for people trying to find truth.³ The problem is that "truth" is at once a straightforward and a very slippery concept.^4,5 In fact, understanding the "truth concept" could be regarded as at the core of philosophy, and its analysis has perplexed philosophers for millennia. What emerges from this debate is that whatever truth is, it is almost certainly not what people believe it to be; and neither is it that which motivates the arguments and actions of social media groups around fact-checking.

According to Meta, its process of fact-checking involves online statements that trigger some sort of suspicion being tested by independent fact-checkers against “sources…public data, authenticating images and videos and more.”⁶ This system was launched in 2016 to verify claims made on social media and remove misleading ones. However, Meta now thinks this system is problematic,¹ claiming fact-checkers have acquired too much authority in determining truth: “When we launched our independent fact checking program in 2016, we were very clear that we didn’t want to be the arbiters of truth.”¹ Instead, Meta proposes removing the arbitration of truth from a small group and giving that power to the community of their social media users (community notes)—who can comment on the likely veracity of posts, and “…decide when posts are potentially misleading and need more context, and people across a diverse range of perspectives decide what sort of context is helpful for other users to see.”¹

These actions suggest that Meta’s concept of truth is psychologised—one that attacks the view that truth is what is real, and regards truth as determined by social and personal consderations.⁴ If you favour this latter view, then you will feel empowered or enraged by Meta’s decision, depending on your group affiliation. However, if you favour a view that truth is not socially constructed, but involves "perceiving things according to their nature," as suggested by Heraclitus,⁴ or “…harbours something of the sublime. Its pursuit is a noble pursuit…”⁷ then you may feel that none of this is getting at the real issue. For some philosophers, whether a small group of experts or a larger community best determines truth is to misunderstand the nature of truth. For them, truth does not depend on agreement. It is independent of such psychological considerations, and it can be known by one person, standing alone, or even nobody.

Given there are two (at least) views of truth, why is there a fair degree of outrage when things are done to alter the way statements are checked for veracity against different social norms (experts versus users)? Why do people not just shrug and say: "check consensus any way you want; it will not get at the truth"? The answer may be found in studies suggesting that most people adhere to a consensus view of truth (irrespective of what they say).⁸ In an experimental study,⁸ people were asked to judge the truth or falsity of social media statements. They accepted information as true when most things in that environment were taken to be true, when they matched their memory, and when it was easy to understand and process the statement under scrutiny.

A Thought Experiment

The question is this: Are people correct in these consensus views of truth? A thought experiment may help to explore this. Imagine a group of people enter a room on the 20th floor of a building, through its single door located toward the edge of one wall. In the room, they settle in for a long stay and debate the nature of the door; perhaps they are fans of Umberto Eco. After much debate, they all agree that they believe the door not to be at the edge of the wall but in the middle of the wall. After all, there is a slight difference of colour in that area from the rest of the wall. They all start to use the word "door" consistently to refer to this patch of grey in the middle of the wall, and they freely discuss how, one day, they are going to go in and out of this "door."

What can we say about their consensually agreed use of the word "door"? Has their consensus constructed a truthful doorway? The "community notes" system would imply it has. Moreover, if they stay in there for a really long time, their memories may imply this as well. If they write down their views, then somebody could even fact-check their statement against their public records.

To test this, let’s say a fire breaks out and is rapidly engulfing the building. Our group needs to get out, and somebody shouts: "Quick, to the door!" They have a choice to make. If a realist-correspondence theory of truth is correct, they will run to the real door (the one that matched their established reality). If a psychologised-consensus view of truth is correct, they will run at the wall, where they point to when they say "door" (that matches their established and checked consensus). I’m fairly sure of what would happen with our group (not so sure about their descendants), but let’s just say the fire brigade arrives, brings a very long ladder, smashes the windows, and rescues our intrepid group before they need to make that decision.

If you think the group would have run at the real door, then you believe in a realist-correspondence view of truth, such as that held by philosophers like Davidson⁵: “…we are justified in asserting a sentence…only if we believe the sentence we use to make the assertion is true; and what ultimately ties language to the world is that the conditions that typically cause us to hold sentences true constitute the truth conditions, and hence the meanings, of our sentences.” That is, describing the truth must make a difference to our behaviour in the world. Now, it may be that most people do this, and such statements would pass a consensus test, but that is not guaranteed. The only way to really know if it’s true is to test the statement against reality. If this is the case, then why do we care whether Meta drops one consensus-based way of checking the truth of posts in favour of another? In either case, they are not dealing in truth.