Seeking sense is very easy to describe and very hard to actually do in practice. The basis of this technique is a claim is that the things that go on in other people's heads make sense from their perspective and that it's valuable to try to figure out what's happening in their brains. (Other people know things too!)

We can divide this technique into two parts: the skill and the attitude.

A seeking sense attitude is approaching other brains in a way that makes you curious about how they work. When you observe someone doing something that ranks very low according to your values, instead of dismissing them as bad, you find yourself very curious as to what process is making them take that action. This notion is partially captured by words like "empathy" or "compassion". To me, it feels like the seeking sense attitude means I have to treat other people like human beings instead of just action/reaction APIs.

The seeking sense skill is the ability to actually resolve the curiosity generated by the seeking sense attitude. In some sense, this is the rationality version of having social skills. Having this ability allows you to observe people's actions and infer things about their values, how they think, what they know about the world, etc.

Both of these things are related to boggling at other people's brains.

Seeking Sense

The first thing that you must accept in order to seek sense properly is the claim that minds actually make sense. This sounds extremely reasonable, but often, I find that it's difficult to accept that that a given mind might make sense. For example, people who seem to be taking actions that aren't optimizing for anything confuse me a lot; it takes a strong perspective shift for me to get into the mode where I think what they're doing might make sense.

Something that might help is realizing that your actions might not make sense from other people's perspectives. When I imagine these people, I find myself compelled to say "Well, actually, my actions make perfect sense because of X, Y and Z." I imagine myself doing this because I think I have a decent understanding of why I do the things I do. In many cases, however, people don't quite understand their own minds well enough to offer clear explanations. This makes seeking sense difficult because you can’t just ask them to explain. Nevertheless, you can imagine what sorts of explanations that they might offer you if they had such understanding.

The second thing that you must accept is that it's possible for you to figure out what's happening in other people's brains. This claim is less difficult to accept because most brains are pretty similar to other brains along most axes. I'm sure that most of you have had moments in your life where you didn't realize why people were acting the way that they were until some key insight into human behavior. My friend recently realized that people can have emotions but not be consciously aware that they're having those emotions, which means a lot more actions now make sense to them.

Given the above claims, the way to seek sense is to actually try and seek sense. This is much harder than it actually seems. One of the theories for why intelligence exists is because we needed to model other agents. This theory doesn't seem complete, but it does ring true to me. When I try to think about other people, my instinct is to think of them as action/reaction APIs and ask questions like "how can I get you to do what I want?" and "what use are you for me?" By default, when we look at people, we see schemers and plotters and planners that are trying to compete with us for limited resources; there's no room for seeking sense.

Nate Soares offers a possible way to shift your perspective:

If you back a puppy into a corner and frighten it, and it snaps at you, it's easy to feel a wave of compassion rather than hatred.

But when a human snaps at you, the social machinery engages. It's easy to get stuck inside the interaction. When a human is backed into a corner and lashes out, we tend to lash back.

If the person who just acted in that way didn't trigger all of your social manipulation brain circuits, would you be more curious as to why they were acting that way? If you thought of them as torn from their home on the savanna and thrust into this bewildering world, would you feel more inclined to think of them as a person? Would you want to try and understand? I would.


I initially didn’t have concrete steps in this post. I am reluctant to give them because I want to avoid the failure mode of “I did the steps but it still doesn’t make sense, which means that they’re stupid.” Please keep in mind that this is only one possible method to seek sense and not even one that I use very often. Most of the time I seek sense, it’s not done in any explicit way. Being very curious about what’s going on in someone’s head is sufficient much of the time.

Nevertheless, here are the steps: 0. Observe some confusing behavior.

  1. Guess what values/knowledge the other person has.
  2. Model yourself with those values/knowledge and check if you would perform the same action.
    • If yes, then you’ve sought sense!
    • If not, then try to refine your guess and return to (1).


A common failure mode is to attempt to seek sense and accept a bad explanation for what's happening. An exaggerated example of this happening is someone observing someone else acting in a seeming contradictory way and "seeking sense" by thinking "oh, they're just stupid. That's why they're acting in that way. That makes sense." When you're trying to figure out what's going on in other people's brains, it has to make sense. Beware of fake explanations. If you can "seek sense" in a way that explains every possible behavior, then you're not actually seeking sense.

Scene: friend has a technical interview tomorrow. They are doing simple practice problems.

me: Why are you doing simple problems? If you expect to be able to solve the problems, you aren’t learning anything. You should be doing difficult problems instead. friend: The interview is tomorrow, so I can’t learn anything in time. My performance is related to my confidence which is controlled by how many problems I can do successfully.

The failure here was accepting the first explanation that I thought of, which was that my friend was acting suboptimally. If I had paused and thought for a bit longer, I would have realized that I had observed them doing difficult exercises in the past. I also might have realized that learning anything significant the day before an interview was unrealistic. Instead, I observed a behavior that I ranked as low-value and assumed that my friend was acting that way because they couldn’t think of a higher value activity. In reality, they ranked the activity as high-value. Luckily, my friend knew enough about themself to know why they were acting the way that they were.

You can also seek sense with objects instead of people. This is very similar to Dennett's intentional stance. You pick an object, assume that it was designed in a way that makes sense, and use that assumption to infer various properties. You can think of this as seeking sense in the designer of the object (or the team of designers).

Here's another example of when I failed to seek sense (why are all my examples of failures?). I have a pepper grinder. The pepper-grindy bit at the bottom can be unscrewed. When my pepper grinder ran out of peppercorn, I naturally tried to unscrew the pepper-grindy bit to put more peppercorn in. However, the opening was very small and only allowed me to put in about two peppercorns at the same time. I reasoned that the pepper grinder was just poorly designed and spent a few minutes every few months struggling to refill my pepper grinder for the next few years.

One day, I was trying to refill my pepper grinder and I dropped it. When it hit the ground, the top of the pepper grinder came off, revealing a very large opening obviously intended to allow for easy peppercorn refilling. In this case, I failed to realize that usually, products are well designed (although sometimes they are). I (wrongly) assumed that I knew how to use the pepper grinder properly, which prevented me from even trying to find other ways to use it.

Application: Pedagogical Content Knowledge

Pedagogical content knowledge (PCK) is a specific application of seeking sense to teaching. When you're teaching a concept, you have two types of knowledge: general knowledge about how to teach and specific knowledge about your concept. We call the former pedagogical knowledge and the second content knowledge. The intersection of these two types of knowledge is pedagogical content knowledge. Pedagogical content knowledge is knowledge about how to teach the specific thing you're trying to teach.

For example, if I'm trying to teach seeking sense, I have pedagogical knowledge that providing examples is helpful and content knowledge about what seeking sense is. I also (might) have pedagogical content knowledge about the specific ways attempt to teach seeking sense might go wrong, i.e. people might falsely assume that they're seeking sense when they're just accepting bad explanations.

As another example, if I'm trying to teach a low level programming language, I might have pedagogical content knowledge about the various ways that students confuse a pointer-to-the-thing with the thing itself. Given this knowledge, I might spend extra time stressing the difference between the two things, providing examples of why they're different, drawing clear diagrams explaining the difference, providing metaphors/analogies, etc.

In my view, having PCK differentiates good teachers from bad teachers. Teachers that have PCK explain concepts in a way that makes them understandable, anticipating mistakes and pre-empting possible misunderstandings.

Finding PCK amounts to seeking sense in the mind of a student that's trying to learn the material. It involves looking at the mistakes they make, what confuses them, and what sorts of questions they have and trying to construct the model that they have of the material. If you can figure out how your student is modeling the material, you can figure out how their model is different from your model and attempt to guide them to the correct model.

In my experience, finding PCK can be done by finding the minimal example that doesn't make sense. Most larger high-level misunderstanding comes from a small low-level misunderstanding. As soon as your student is confused by a high-level example, construct a number of low level examples that are each related to concepts used in the high-level example. The minimal example that doesn't make sense to the student is an extremely helpful pointer as to how they're thinking of the material.

Importantly, you can try to find PCK with yourself. Often times, when I try and learn a subject, I am confused about a lot of stuff but I don't really know how I'm thinking about the thing I'm trying to learn, so I don't really know why I'm confused. Sometimes, if I figure out the parts I don't understand and the mistakes I'm making, I can find minimal examples and guess how I'm thinking about the subject. I can then go back to the textbook and try and figure out where my thinking differs from the "correct" line of thought (this is the hard part).


Pick a behavior that you see people exhibiting that doesn't make sense to you and try to seek sense in the behavior. Maybe something that your friend/coworker/neighbor/parent does confuses you. Maybe there's an action that you've observed that you've explained away as "weird" or "stupid".

Preferably, there will be reasons to suspect that the sense behind the behavior is simple.

New Comment
6 comments, sorted by Click to highlight new comments since:

The paradox of sense is:

If you know what the purpose is, you can tell if something is doing it badly, or well.

If you think something is doing (its purpose) badly, you might be missing the purpose (or the instruction manual).

(Though if something is good for multiple things, seeing one thing it does well and stopping there might prevent you from seeing its other purposes.)

One day, I was trying to refill my pepper grinder and I dropped it. When it hit the ground, the top of the pepper grinder came off, revealing a very large opening obviously intended to allow for easy peppercorn refilling. In this case, I failed to realize that usually, products are well designed (although sometimes they are). I (wrongly) assumed that I knew how to use the pepper grinder properly, which prevented me from even trying to find other ways to use it.

Weird advice: If something is badly made, consider what would happen if you dropped it. Is it easily/conveniently replaceable?

I'm not sure I understand what you're saying. I'm not advising people to drop their items in an attempt to discover new uses for them, I'm just describing a time when I accidentally dropped something and discovered that I was using it wrong.

I think part of what I'm saying is that your priors for things working properly should be higher and I should have been more surprised at the difficulty of refilling the pepper grinder. This should have prompted me to search harder for a way to use it more effectively.

I'm not advising people to drop their items in an attempt to discover new uses for them

Yes, you are not.

This should have prompted me to search harder for a way to use it more effectively.

I think 'dropping things' is one, perhaps inefficient, way of doing that.

And it makes a good metaphor. If you try things differently, or try new things, they might not work the first time. (Or ever - we remember the Apollo missions, and the Wright Brothers because they succeeded.)

Dropping items in an attempt to discover new uses for them, drawn out over 27 lines:

If you take something apart, you might learn.

But it might break.

So if you dropped it and it broke would that be really inconvenient, or easily replaced?

If something falls it might break.

There might be an opportunity to learn.

To put the pieces back together well.

But there is risk in things falling.

And breaking.

Sometimes they break forever.[1]

There is less risk in taking things apart.

But we don't do it very often.

And sometimes we stop before finishing, because we're afraid of breaking things.[2]

But if something is easily replaced

And we're not afraid of breaking it

Then we might learn something by taking it apart.

If it breaks it breaks.

If we learned something, we learned something.

If we learn a better way of doing or making things, we learn a better way of doing or making things.

Is a broken thing too high a price to pay?

For knowledge?

For a chance to learn a better way?[3]

[1] You might have to learn, how to make glue (red link).

[2] If this isn't you, then this...isn't you.

[3] Even if it takes more than one thing broken?

Until you find a way

to put it back together.

Until you find, another way/how, to use it.

> The first thing that you must accept in order to seek sense properly is the claim that minds actually make sense

This is somewhat weird to me. Since Kahneman & Tverski, we know that system 2 is mostly good at rationalizing the actions taken by system 1, to create a self-coherent narrative. Not only thus minds generally don't make any sense, my minds in general lacks any sense. I'm here just because my system 1 is well adjusted to this modern environment, I don't *need* to make any sense.

From this perspective, "making sense" appears to be a tiring and pointless exercise...

System 1 doesn't make sense?

Many years ago, I’m 46 now, so this would have been 26-28 years ago, I was introduced to the concept referred to as empathic assertion. It was a course in cognitive psychology that I was taking and the idea introduced was placing your empathic, or sensing/feeling self into the “shoes“ of another that was experiencing some type or sort of emotional experience. What you’re writing here reminds me of that practice. To be transparent, I adopted this as a life skill that I still use to this day.