Almost Everything, Somebody Told You

Take an honest inventory of what you know.

The age of the earth. The existence of Antarctica. Your own date of birth. That the heart pumps blood. What is in the food you ate today. Whether a medicine works. The existence of countries you have not visited, which is most of them, and of historical events you did not witness, which is all of them.

Now ask how many of those you verified yourself.

Essentially none. The overwhelming majority of what any person knows arrived because somebody told them, and the theory of knowledge spent most of its history discussing the small remainder, the bit you checked with your own eyes or worked out with your own reasoning.

That is a strange allocation of attention, and it has a cause. Philosophy inherited a picture in which the ideal knower is a single mind working alone, and testimony looked like a concession to human weakness rather than a route to knowledge in its own right. Descartes sitting by his fire, doubting everything, is the image the subject was built around.

But consider what that ideal actually delivers. A person who believed nothing they had not personally established would know their immediate surroundings and some arithmetic. They would not know their own age.

So trust cannot be the exception. It has to be the default, and the real question is not whether to extend it but when to withhold it.

That question has answers, and they are more interesting than the usual advice to think for yourself.

The first is independence. A claim repeated by a hundred people who all heard it from one person has one source, not a hundred. This is the single most common error in assessing testimony, because volume is very easy to mistake for confirmation, and the internet is a machine for multiplying volume without adding sources.

The second is whether the claim passed through anything that could have caught it. A finding that survived people trying to knock it down is in a different position from one that nobody checked, even if both are asserted with equal confidence. The useful question is not how sure the speaker sounds but what would have happened if they had been wrong.

The third is whether the person telling you is in a position to know. This sounds obvious and is constantly ignored, because confidence and authority travel across subjects while expertise does not.

Which brings us to the part of this chapter that is not only about philosophy.

If most of what anybody knows is testimony, then the health of the institutions that produce and check testimony is not a political matter that happens to sit nearby. It is directly a question about how much a person can know.

A society whose testimony has degraded is not a society with the same knowledge and worse manners. It knows less. The individuals in it are not compensated by being more sceptical, because scepticism does not generate findings; it only declines to accept them.

That is an uncomfortable conclusion for a tradition built around the lone thinker, and it follows from taking an honest inventory of what is actually in your head.