The word “Trust” was never quite operationalized as well as it should have been in society, and as a result it can now be used to mean two rather different things.

The first form trust takes is probably the most commonly understood use of the word; expecting someone to behave in a way that’s cooperative or fair. If you trust someone enough, you may enter into a business partnership with them or let them borrow your belongings or vouch for them to friends or colleagues. This trust can be broken, of course, if they start to act in ways other than what you expect them to, particularly if they start to defect from agreements. It is, ultimately, about how well you can model their ability to act prosocially.

The second form trust takes is much rarer, and yet somehow feels to me more like the “true” meaning of the word. It’s a level of trust that’s related to your confidence in someone’s character, sometimes despite their actions. It’s not about predicting what they’ll do in any given situation, but rather predicting the arc that their actions will take over a long enough timeline; trusting them, essentially, to error correct.

This may seem like it has the same outcomes, like if you trust them enough in this way you’d still be okay with lending them something, but it’s far less reliant on game theory or incentives, and far more about what you believe about what kind of person they are. In the first case, if the person you trust does not give back what you lent them, your trust is broken. In the second case, if they do not give back what you lent them, your trust endures, because your expectation is that their character is one who had a good reason not to give it back. This doesn’t require a resolution; it’s baked into the decision to lend them the thing itself, as you’d expect yourself not to regret lending it to them if you had all available future information, and are thus okay with not having that information.

That’s why, in this second sense, “Trust” really only has meaning if it’s applicable to situations where you might normally trust someone less or be unsure of them. If you can always know what someone does and why, your trust of them lacks the real power of the second definition. It’s only when someone is able to act without your knowledge, or acts in ways that you don’t understand, or even that seem like they harm you, that your “true” trust in them is tested, and either justified or not.

Because it can be unjustified. People can trust others in this “true” sense and still be wrong, and be hurt as a result. I think this is why it’s such a rare form of trust, in the end; it’s a more vulnerable stance to take, the same way an expression of love is different from an explicit commitment.