2 min read
2 min read

16

Followup to: Tonic Judo

Related to: Correspondence Bias

Imagine that someone you know has a reaction that you consider disproportionate to the severity of the event that caused it. If your friend loses their comb, and they get weirdly angry about it, and you persuade them into calming down with rational argument, and then it happens again, say, many months later, and they get just as angry as they did the first time, is that person unteachable? Is it a waste of your time to try to persuade them using rationality?

I think a lot of people would have an expectation that the friend would not have another outburst, and that when the friend had another outburst, that expectation would be violated.

And for some reason, at this turn, it seems like a lot of people think, "I tried to teach this person once, and it didn't work. They're the kind of person who can't be persuaded. I should direct my efforts elsewhere." Maybe you even make it look more 'rational' by name-dropping expected utility.

Or maybe it doesn't feel like stubbornness; maybe it feels like they just forgot. Like they were pretending to listen when they looked like they were listening to your arguments, but really they were just waiting for you to finish talking.

That does happen sometimes, if you fail to emotionally engage someone or if you're hanging out with all the wrong kinds of people.

But most of the time, when you're dealing with the majority of the human race, with all of the people who care about how they behave, the right way to go is to realize that a violation of expectations is a sign that your model is wrong.

You made your first rational argument with the implicit expectation that it would prevent all future outbursts over combs. But it happens again. You shouldn't stop at your first attempt. It may be that circumstances are different this time and an outburst is warranted, or it may be that your friend is not in a state in which your previous arguments are at the level of their attention. Or maybe they feel righteous anger and you need to get them to have less self-confidence and more confidence in you, and maybe you need to encourage them to control that in the future, instead of only the previous object-level impulse.

The point is, you expected your first argument to generalize more than it actually did. People often respond to situations like this as though the fact that their first attempt to instill a very general behavior in another person is strong evidence that the person can never be made to instill that general behavior. It's only strong evidence that your first attempt to instill a general behavior was less successful than you expected it to be.

The idea is to keep up your rational arguments, to give them enough feedback to actually learn the complicated thing that you're trying to teach them. From the fact that you see that your arguments generalize in certain situations, it does not follow that you have successfully given others the ability to see the generalizations that you can see.

(Content note: Inspired by this comment by user:jimmy. Highly recommended reading.)

New Comment


18 comments, sorted by Click to highlight new comments since:

I've heard that as a behavior is being extinguished, it becomes less common, but it's just as intense when it happens. That's something else to check.

This makes intuitive sense, anyway: someone who is about to smoke (what turns out to be) their last cigarette before they finally succeed in quitting, does not therefore do a half-assed job of lighting it.

[-][anonymous]00

True at least in lab conditions (it's harder to measure force in real world conditions), but this isn't an extinction process. The antecedent of losing something is still present; the consequence is less clear but there's no indication it has changed. Counseling could be seen then as positive reinforcement of an alternative, or as positive punishment of getting angry; assuming it causes a change in behavior.

Interesting, I'll take a look around. Google keywords you would use or any links that you might be able to dredge up from a shallow search would be appreciated.

Imagine that someone you know has a reaction that you consider disproportionate to the severity of the event that caused it. If your friend loses their comb, and they get weirdly angry about it, and you persuade them into calming down with rational argument, and then it happens again, say, many months later, and they get just as angry as they did the first time, is that person unteachable? Is it a waste of your time to try to persuade them using rationality?

There are two ways you can interpret rationality in that sentence:
(1) Appeal to explicit reasoning by providing objectively true arguments why being angry doesn't make sense.
(2) Appeal to effective actions.

Don't confuse the two. They are very different in this case. The kind of rational persuasion that Jimmy with his hypnosis background talks isn't simply just providing objectively true information.

At the beginning of the last LW Europe Community Camp we went out bouldering. There was a girl who at the beginning had a phobia of heights that prevented her from going on a slackline. Two hours later she didn't have the phobia anymore, and found that highly surprising. She also had a public speaking phobia and gave half a year later a talk in front of 1000 people.

As far as amnesia goes, amnesia isn't a problem. In hypnosis amensia sometimes even get's consciously added. If you actually remove the emotional attachment that produced the problem it doesn't matter whether they remember the conversation the next time the friend loses their comb.

Imagine that someone you know has a reaction that you consider disproportionate to the severity of the event that caused it.

What you describe (someone's button's being pushed) is commonly known as a trigger. What they are experiencing might even be related to PTSD, though probably not in your example with the misplaced comb.

The idea is to keep up your rational arguments, to give them enough feedback

This may work in isolated cases, but most of the overreaction you describe is subconscious. Logic does not work well on the subconscious mind, if at all.

to actually learn the complicated thing that you're trying to teach them.

This often comes across as condescending and may easily cause resentment.

Logic does not work well on the subconscious mind, if at all.

No, but you are talking to a whole person who, hopefully, has other brain functions besides subconsciousness.

What you describe (someone's button's being pushed) is commonly known as a trigger. What they are experiencing might even be related to PTSD, though probably not in your example with the misplaced comb.

Yeah, I have tried to keep in mind that there might be some time in my friend's past where it was useful for him to freak out like that.

This may work in isolated cases, but most of the overreaction you describe is subconscious. Logic does not work well on the subconscious mind, if at all.

I think this is a little too simplistic. You can bring things to the level of someone's attention that wouldn't have been noticed without guidance, and by repeatedly relating immediate contexts to previous contexts that included rational arguments, people can learn how to do new things just like they have their entire lives. The fact that we call them rationality techniques doesn't make them different from all other skills in any special way. They're just metacognitive skills in particular.

This often comes across as condescending and may easily cause resentment.

I don't use that word when I'm talking to someone. I did think, after the fact, that all of this talk of 'instilling behaviors' and 'teaching' would sound a little too Orwellian, but hopefully everyone can see that this is not that, and that this is not very different from how people manage social expectations all of the time. You're just explicitly optimizing them now.

Why not try buying them a dozen combs for $10? Then there can be combs all over the apartment and it doesn't matter if one gets lost and nobody needs to get mad or frustrated.

Meta point: Don't reject solutions on the object level out of hand. Remember the hair dryer incident.

Maybe you even make it [wanting to direct your efforts elsewhere] look more 'rational' by name-dropping expected utility.

This is an interesting point. It has often been my experience that the proposition "helping friends work out mental bugs is a good use of time in expectation" feels true in the abstract more often than it feels true in the sum of all particular cases.

I'm currently inclined to say that helping friends work out mental bugs actually is worth it in general, though I am currently in abstract-thinking mode as I'm saying this. ;)

[-][anonymous]00

If this happens, I would recommend looking into whether there were certain times where he lost something and didn't get angry, and if so why, or looking into why he's losing items. What are the shared features of the lost items? Behavior change usually doesn't occur by directly targeting the thing you want to change, but by indirectly targeting the behavior; by looking at how and why the alternatives occur. That is what you were doing with counseling another way to react to the situation. If the behavior could change by directly targeting the behavior, the problem likely never would have become prevalent enough to come to your attention.