People have an annoying tendency to hear the word “rationalism” and think “Spock”, despite direct exhortation against that exact interpretation. But I don’t know of any source directly describing a stance toward emotions which rationalists-as-a-group typically do endorse. The goal of this post is to explain such a stance. It’s roughly the concept of hangriness, but generalized to other emotions.

That means this post is trying to do two things at once:

Illustrate a certain stance toward emotions, which I definitely take and which I think many people around me also often take. (Most of the post will focus on this part.)
Claim that the stance in question is fairly canonical or standard for rationalists-as-a-group, modulo disclaimers about rationalists never agreeing on anything.

Many people will no doubt disagree that the stance I describe is roughly-canonical among rationalists, and that’s a ~~useful~~ valid thing to argue about in the comments in proportion to how well you actually know many rationalists.

Central Example: “Hangry”

When we’re hangry, it feels like people around us are doing stupid, inconsiderate, or otherwise bad things. It feels like we’re justifiably angry about those things. But then we eat, and suddenly our previous anger doesn’t feel so justified any more.

When we’re hangry, our anger is importantly wrong, or false in some sense. The feelings are telling our brain that other people are doing stupid, inconsiderate, or otherwise egregious things. And later, on reflection, we will realize that our feelings were largely wrong about that; the feelings were not really justified by the supposed wrongdoings.

But the correct response is not to dismiss or ignore the feelings! Even if the feelings “tell us false things” in some sense, those feelings still result from an important unmet need: we need food! The correct response isn’t to ignore or dismiss the anger, the correct response is to realize that the anger is mostly caused by hunger, and to go eat.

The word “hangry” conveys this whole idea in two syllables. And crucially, the existence of “hangry” as a word normalizes the phenomenon - more on that later.

I consider the word “hangry” to be one of the main ways in which mainstream society has become more sane in the past ~10 years. In a single word, it perfectly captures the stance toward emotions which I want to describe. We just need to generalize hangriness to other emotions.

The Generalized Hangriness Stance

The stance itself involves three main pieces:

Emotions make claims, which can be true or false.
Even false claims convey useful information, just not the information they say they convey.
So, the core move is to ask “Setting aside what this emotion claims, what information does it actually give me?” (which, to be clear, will sometimes-but-not-consistently match what the emotion claims).

Emotions Make Claims, And Their Claims Can Be True Or False

Words have semantics. If someone tells me “there’s a bathroom down the hall around the corner”, then when I walk down the hall and turn the corner, I expect to see a bathroom. A physical bathroom being in that physical spot is the main semantic claim of the words.

Likewise, emotions have semantics; they claim things. Anger might claim to me that it was stupid or inconsiderate for someone to text me repeatedly while I’m trying to work. Excitement might claim to me that an upcoming show will be really fun. Longing might claim to young me “if only I could leave school in the middle of the day to go get ice cream, I wouldn’t feel so trapped”. Satisfaction might claim to me that my code right now is working properly, it’s doing what I wanted.

As with words, those semantic claims can be true or false.

If someone claims to me that there’s a bathroom down the hall around the corner, and then I go down the hall and around the corner and there’s no bathroom, I update that their claim was probably false. (Even more so if it turns out there is no corner, or possibly even no hallway.) If I go down the hall and around the corner and find a bathroom, then the claim was true.

If my anger claims to me that it was stupid or inconsiderate for someone to text me repeatedly while I’m trying to work, but on reflection I realize that I didn’t indicate I was busy and can’t reasonably expect them to guess I was busy, I update that my anger’s claim was probably false. If on reflection I have told the person many times before that texts during work hours are costly to me, then I update that my anger’s claim was probably true.

If my excitement claims to me that an upcoming show will be really fun, and the show turns out to be boring, then the claim was false. If the show turns out to be, say, the annual panto at the Palladium, then the claim was very conclusively true.

If my longing claims to young me “if only I could leave school in the middle of the day to go get ice cream, I wouldn’t feel so trapped”, and upon growing older and having the freedom to go get ice cream in the middle of the day I still feel trapped, I update that my longing’s claim was probably false. In fact I do now have the freedom to get ice cream in the middle of the day, and I generally do not feel trapped, so that’s an update toward my longing’s claim being true.

If my satisfaction claims to me that my code right now is working properly, and it turns out that an LLM simply overwrote my test code to always pass, then my satisfaction’s claim is false. If it turns out that my code is indeed working properly, then my satisfaction’s claim is true.

In general, if you want to know what an emotion is claiming, just imagine that the emotion is a person or cute animal who can talk, and ask what they say to you.

False Claims Still Contain Useful Information (It’s Just Not What They Claim)

Let’s say I feel angry, so I imagine that my anger is a character named Angie and I ask them what’s up. And Angie starts off on a rant about how this shitty software library has terrible documentation and the API just isn’t doing what I expected and I’ve been at this for three fucking hours and goddammit I’m just so tired of this shit.

So, ok, Angie claims to be angry about the shitty software library. Fair enough, most software libraries are in fact hot trash. But c’mon, Angie, usually we’re not this worked up about it. What’s really going on here? And Angie pauses for a moment and is like “Man, I am just so tired.”. Perhaps what is really needed is… a break? Perhaps a nap? Perhaps a snack or some salt (both of which often alleviate tiredness)?

In a case like this, my anger is making claims about the quality of a software library. And those claims are… probably somewhat exaggerated in salience, even if not entirely false. But even insofar as the claims themselves are false, they still convey useful information. The anger may be wrong about the quality of the software library, but it still contains useful information: I’m tired. As a rough general rule, strong emotions are strong because some part of me is trying to tell me something it thinks is important… just not necessarily the thing the emotion claims.

“Pretend the emotion is a person or cute animal who can talk” is a pretty great trick. Not just for checking what they say, but for checking what they don’t say. See, lots of people have good enough social instincts to ask “Is that what’s really bothering you?” when someone else is worked up, but it’s a harder skill to pose that question to oneself. Picturing the emotion as a person or animal triggers that external perspective, makes it easier to notice that maybe the emotion is bothered by something other than what it’s saying.

But you can also just ask yourself “What’s really generating this emotion? What can I actually guess from it, setting aside the claims the emotion makes?”.

... and once one starts down that path, very often the answer turns out to be "I'm scared of X and this emotion wants to protect me from X". Often X is social disapproval of some sort (ranging from a glare to outright ostracism), or something the person has been burned by in the past. And that's why so many rationalists end up down a rabbit hole of trauma processing, or relational practices meant to make people feel loved and supported, or other borderline woo-ish things. An awful lot of those woo-ish things are optimized to deal with exactly these sorts of emotion generators.

The Generalized Hangriness Stance as Social Tech

Arguably the best thing about the word “hangry” is that its existence normalizes hangriness.

20 years ago, if some definitely-hypothetical person suggested to their hypothetical romantic partner that perhaps the partner was not angry about the thing they were ranting about, but was instead grumpy from being hungry… yeah, uh, that would normally not go over well. The partner would feel like their completely-valid(-feeling) emotions were just being brushed off, with some nonsense about being hungry.

But with the word hangry, it’s a lot easier to say “You seem maybe hangry right now, how about you have something to eat and if you still feel this way after then we can talk about it”. That doesn’t always work; it might still feel like being brushed off if someone’s worked up enough and/or sufficiently terrible at understanding their emotions. (Also people might sometimes in fact try to brush off other peoples’ valid emotions by hypothesizing hangriness, but that’s a trick which only delays things for like 20 minutes if one responds with the obvious test of eating something.) But it works a lot better than trying to convey the same thing before hangriness was normalized as a concept.

Alas, there’s still no word which normalizes this kind of thing more generally.

Telling people in the moment that the things their emotions are telling them seem false, and perhaps their emotions convey some other information… is usually not the right move unless you’re very unusually good at making people feel seen while not not telling them they’re being an idiot. Because, yes, someone ranting due to hangriness is being an idiot, and no, directly telling them they’re being an idiot does not help. Either they need to have already bought into the idea that a very large chunk of most peoples’ emotions claim false things but nonetheless convey useful information, or they’ll need to feel seen before anything else works. And it’s very hard to make such a person feel seen without at least somewhat endorsing whatever idiocy their emotions are claiming.

Among other rationalists, I usually expect that people are on board with the Generalized Hangriness Stance, so it’s usually ok to say something like “Look, I think you feel like X, but I suspect your feeling is in fact coming from Y rather than X. And to be clear I could be wrong here, but I think this should at least be in our hypothesis space. We can look at A, B, C as relevant evidence, and maybe try D, and if that doesn’t work then I’ll update that the feeling probably is coming from Y after all.”. Where, to be clear, the explicit “I could be wrong here” is an extremely load-bearing part of what makes this all work. Another good wording I use frequently is “I’m not sure this is true, but here’s a model of what’s going on…” ideally peppered with frequent reminders that this is a model and I’m not asserting that it’s correct.

Point is, this Stance toward emotions isn’t just individually useful. Arguably most of its value is as social tech. When most people in a space are on board with the Generalized Hangriness Stance, it becomes possible-at-all to point out to people that maybe their emotions are claiming stupid things, without that necessarily coming across as an attack on the person (and triggering defensiveness). And it then also becomes possible to help someone figure out what information their emotions actually convey, and help them with what they actually need (like e.g. eating). Some skill is still required, but it’s much more tractable when there’s common knowledge that people are on board with the Generalized Hangriness Stance.