We take a rather old-fashioned view of what a “fact” is in this text. For us, a fact is a claim about which there is general agreement by people in the know.

Most claims aren’t facts and aren’t intended to be presented as facts. People make claims all the time. I may claim that Mulholland Drive is the best film of the 2000s. You may claim that Spiderman 2 is. People can disagree about these things. These are claims, but they are not facts.

Some types of claims, however, also qualify as “statements of fact.” It is a statement of fact, for example, that Mulholland Drive was directed by David Lynch, and it’s a statement of fact that Sweet Home Alabama starred Reese Witherspoon. Facts don’t have to be physical: it’s a fact that Sweet Home Alabama deals with questions of what is most important in life and that Mulholland Drive investigates how the stories we tell ourselves differ from the reality we inhabit.

Facts can even deal with situations that are hypothetical. We can say that it’s a fact that Sweet Home Alabama would have cost less if it was shot in Serbia instead of in Georgia and Alabama.

For us, a fact is:

• something that is generally not disputed
• by people in a position to know
• by those who can be relied on to accurately tell the truth

That’s it. When we talk about facts, we are usually attempting to get at truth. But the measurement of what qualifies as fact is that it meets those three criteria.

That said, there’s a whole lot to be said about those criteria.

Position to Know: Expertise, Opportunity, and Access

Let’s start with the second one. Who are the “people in a position to know?”

Generally, “a position to know” denotes expertise or opportunity.

Let’s take a car accident as an example. Your car and my car collide on a deserted road. Who is in a position to know what happened? Well, obviously you and I. If we both agree to what happened–say, we both agree that I drifted into the oncoming lane due to a lack of sleep–we can treat that as a fact.

Perhaps someone else disagrees with that. No, says a person who reads the account of the crash in the newspaper, that’s not how it happened at all! Do we suddenly have to start treating this account of the crash as a claim that does not have the status of fact?

It depends. The crucial question is whether this third person is in a position to know. Did they see the crash? Then yes, we have to stop treating our account of the crash as fact. Do they have some deep knowledge of crash forensics that shows the crash is impossible? Are they a crash expert? Well, yes, then perhaps the fact is in dispute, though to override the evidence of our claimed experience, we’d want more than a single expert opinion–we’d want expert consensus.

On the other hand, if they did not see the crash but instead believe that very few crashes happen due to lack of sleep and therefore this cannot be an explanation, then no, we can still treat this as fact, because the people in a position to know are in agreement. The disagreement of a person not in a position to know does not undo that.

With questions involving expertise, “position to know” generally indicates expertise, but even here there is opportunity at work. As an example, consider the recent debate over whether the Russian government was responsible for feeding the Democratic National Convention emails to Wikileaks. To have an informed opinion on this issue, you’d need some expertise in cybersecurity forensics. But just having the expertise is not enough. To evaluate the issue, you’d need access to the systems that were hacked: log files, operating system, etc. Indeed, one of the struggles with that issue was that the people with both access and expertise were not always trusted to tell the truth. (This leads to a situation where we can say it is highly likely the Russian government ordered the distribution of the information to Wikileaks, but we cannot accord it the status of fact).

By the same token, opportunity isn’t always enough. A person may take a photograph, for example, of something they think is a lynx but turns out, when reviewed by experts, to be a cougar. If three people witness an animal dart across the road, and one works in a zoo, we might be inclined to weight the zookeeper’s opinion of what the animal was more heavily.

Because “position to know” is so important to claims of fact, when we look at sources we are always asking ourselves, “What puts this person in a unique position to know?” If the answer is “nothing in particular,” then we find new sources.

One final thing–and it’s perhaps the hardest thing to swallow. In most cases we, the readers, are not in a position to know about specific facts ourselves. Our experience matters, but as readers we tend to vastly overrate its value in ascertaining questions of fact where we have neither expertise or opportunity.

We don’t get a vote on the facts; we get a vote on who is most credible. And this means we usually have to trust someone other than ourselves.

Generally Means Generally

What about this other phrase “generally not disputed”? What does that mean exactly?

It’s difficult to say. Imagine you are interviewing twenty people about a wedding reception. Ten of them say that the wedding cake was chocolate, and ten say, “No that’s wrong, it was yellow cake.” In this case we have a clear dispute, and can’t really treat the type of cake as fact. But say that 19 people said the cake was yellow cake and one said it was chocolate. In this case, we’re likely to treat the type of cake as fact and assume that the twentieth person may have misremembered the event or had some weird ulterior motive for lying about the cake.

What that line is–where something is “generally” not disputed–is a question of much debate. It may vary depending on the importance of the question. For a relatively silly question like cake type, four out of five people may be enough to say the question is settled. For issues of greater importance, “generally” may require a higher percentage. Importantly, though, for questions where a lot of people are in the know, that percentage does not have to be 100%. Why? Because even people in the know make mistakes, and because people in the know may have reasons for not telling the truth.

A good example of this is evolution. Evolution, the process by which organisms evolve into new organisms over time, is fact. It’s fact not because every person on the planet has gone over the evidence individually, but because the people who are in the best position to interpret that evidence (biologists, geologists, zoologists, etc.) are in almost unanimous agreement on this point.

But that does not mean that all scientists are in agreement. In fact, 13 percent of scientists disagree with evolution.

Admittedly, the vast majority of those people are neither biologists nor experts in evolution; they are various chemical engineers, physicists, and medical professionals. However, if you dig deeper, you will even find a biologist or geneticist here or there who disagrees with what most scientists see as one of the foundational truths of biology.

If something is so obviously true, how come there is not 100% agreement?

One answer is that maybe the dissenters have it right, and that certainly has happened before. It was, of course, a fact 600 years ago that the sun went around the earth. So sometimes dissenters are right, and the facts are wrong.

More often, when there is overwhelming agreement and a small amount of dissent, it’s just human fallibility. People looking at the same evidence with the same intelligence and the same authority come to different conclusions. Sometimes the minority have a stake in the outcome that can bias them, or just have a different way of looking at things.

Again, this sort of dissent is good, and is necessary to the progress of science, technology, and culture. But when we are asking whether something is a fact, we are not asking whether something has a 100% chance of being true or whether agreement on it will last for all time. We are asking whether enough people in the know agree on it that we can treat it as a settled question and move forward on it, either to action or more complex claims.

Truthfulness and Fact

It is popularly assumed that the biggest question to ask in media literacy is whether people are lying to you or not. Indeed, bias, the tendency of people’s beliefs, incentives, or financial interests to influence what they promote as “true,” can make some experts or witnesses untrustworthy. If we return to our car crash example: two people crash into one another with different stories on how it happened. It certainly makes sense to examine the motivations each of those persons may have for lying.

At the same time, I’d argue that in fact-checking, it’s more useful to make truthfulness the question you pursue second, not first.

To explain: in any given situation, the people in the best position to know something are generally a small group, since people with both the opportunity to review evidence and the expertise to evaluate it are limited. If you are looking for the fastest way to whittle down who to trust, expertise and opportunity get you there more quickly.

After that point–after you’ve whittled down your trusted circle based on these “position to know” attributes–you may still find disagreement. And at that point, it is useful to ask whether the disagreement is an honest disagreement or is a result of hidden (or not-so-hidden) motivations.

The danger of asking the “truthfulness” question first is that everybody has some bias. So it quickly becomes very easy for a reader to eliminate credible sources because “of course they would say that, wouldn’t they?”

If you ask yourself who would be in a “position to know” first, and go to the truthfulness question second, you’ll be able to validate sources more quickly and more reliably. You’ll also be able to see more clearly any true bias that may be influencing your group of experts.