Thread by @s_r_constantin

I think there are two different concepts, *favor* and *trust*, that are not the same and (maybe) shouldn’t be pushed to be the same.

🧵

You favor someone, or view them favorably, when you are motivated to be their ally, or to be on their side.

When you favor someone, you do nice things for them, defend them against others, and *show* everyone that they’re your ally.

Signs of affection, approval, affiliation, and inclusion in the group are important signals of favor.

You trust someone, or respect them, or grant them credibility, when you believe you can rely on them to do something you value.

When you trust someone’s honesty and financial stability, you might be willing to lend them money. When you trust someone’s expertise/knowledge/intellect/wisdom, you might take their advice.

People can favor someone they don’t trust, for a number of reasons.

1. We can favor someone if we intrinsically value their happiness.

A newborn baby hasn’t earned much trust or respect, but most of us still want to be kind to babies. Especially our own or our friends’.

2. People tend to be more favorably disposed to familiar people. Neighbors, friends-of-friends, people whose internet posts we see regularly, etc.

3. You might favor someone in order to get *their* favor, because you want access to their resources or because you fear them harming you.

And you can trust someone without favoring them. You can have a lot of respect for someone’s skills or virtues but still not want to be their ally. Their agenda might seem bad to you; or you may not like their other allies; or you might just not subjectively like them.

Where this gets confused is in discussions of merit, desert, or “goodness.”

In practice, calling someone “good” or “worthy” can be *either* a statement of favor or of trust, and the ambiguity causes a lot of problems.

People can say “I think he’s good” or “I don’t think he’s malicious” or even “I think he’s right” or “I think he’s being honest” when what they really mean is “I’m on his side; let’s not hurt him.”

But people can also use those statements to mean “I can rely on him to do the right thing” or “I would bet on his statements being accurate”.

This leads to people talking past each other.

It gets especially weird in the case of behavior that is both harmful/unjust and common.

Suppose you live in a slave-owning society. Lots of people own slaves and even more support or tolerate slavery.

Are slavers “bad people”?

Well, slavery is very bad and they are intentionally practicing it, so yes, in the “trust” sense, these are people who habitually do bad things. Obviously.

On the “favor” side of the equation, though, if you lived in this society, depending on who you happened to be born, you might genuinely *like* a lot of slavers; they might be your family and friends, or even yourself. Even if not, they are humans and neighbors.

You probably wouldn’t *want* to kill all the slavers if they were 20% of the population; and the reasons have to do with practicality and basic human sympathy. It probably wouldn’t work, and even if it did, it would probably gross you out.

A wrong doesn’t become less wrong if it’s practiced by more people; but severe punishment *does* become less practical and desirable for common wrongs than rare ones.

When people talk about justice, we tend to want to limit *favoritism*, or showing more favor to your ingroup independently of objective criteria. Favoritism is supposed to be inappropriate in courts of law, for instance. This seems good to me at least in specific contexts.

Societies that run on favoritism tend to have problems with corruption and clan warfare.

You definitely want some things — like a clean city water supply — to be functional, and for decisions to be judged on effectiveness, not based on who the Big Man wants to favor.

so sometimes you clearly want to make decisions on a “trust” basis. Assign people to jobs if and only if you expect them to do well. Take advice or rely on info from people if and only if you expect that to work out well.

And to some extent, favor and trust must be linked; you want to reward trustworthy people enough, relative to untrustworthy people, to incentivize more people to become trustworthy.

OTOH, I’m not at all convinced that *every* choice a person makes about whom to be nice to, or whom to affiliate with, should be a reward for trustworthiness/virtue.

Sometimes you just like someone! Or don’t!

Human motives are more complicated than that. The person you like the best doesn’t have to be “the best in the world” at any metric you can clearly articulate.

There are lots of people I like, enjoy having around, would defend against social attack — but probably wouldn’t lend money to, or let babysit my kid.

I think people vary in how “tightly coupled” favor and trust are for them.

There are people who feel uncomfortable/unsafe/disgusted around people they don’t trust/respect on serious stuff.

I think I’m more “loosely coupled” than most.

“Could I count on this person?” doesn’t feature that strongly in how much the idea of interacting with them appeals to me.

From my perspective it looks like the world is full of people yelling “trust me! I’m offended you don’t trust me!” when they obviously haven’t earned trust;

and other people yelling “ugh! Stop being nice to the undeserving!” in situations when being nice is pleasant and prudent whether the recipients are “deserving” or not.

“These people seem to be HAVING FUN WITH EACH OTHER even though they’re not that smart/competent/attractive! Shame on them!” is one trope that annoys me.

But there’s also a trope of “you can’t just evaluate people by objective standards! At least not if they’re in my ingroup!” that’s obviously messed up.

It can be clarifying to distinguish between how much you like someone and an objective evaluation of what they do.

I also think there’s a confusion about wanting to be liked, included socially, shown affection, etc. vs wanting to be evaluated positively.

I want to be liked. I don’t care that much if you think I’m an idiot so long as you think I’m a lovable idiot.

I also, overall, would rather be right than wrong, but I don’t care that much independently about being *perceived* to be wrong.

And I think this is actually saner than caring directly about being “perceived well.”

Affection and social bonds are directly rewarding and useful. Being trusted is only instrumentally useful sometimes.

The capacity of a society to produce/preserve nice things depends on accurate record keeping about credit/trust/credibility. You don’t actually *want* those judgments falsified, even “in your favor.”

People liking you or being nice to you, by contrast, doesn’t have to endanger all the getting-stuff-done machinery; it can *trade against* other uses of resources but it doesn’t undermine them in the same way.

So that’s a downside of perfectly aligning “favor” and “trust.”

If the only way to get people to be nice to you is to be trusted, then everyone has an incentive to deceive about how trustworthy they are, and we lose the ability to coordinate/get things done/not die of plague.

The way I’ve been thinking of this is that there are multiple things you can “spend” your “favor budget” on, and incentivizing trustworthiness is only one of the things people want to “buy”.