[Cross-posted from Grand, Unified, Crazy.]

“You can never make the same mistake twice because the second time you make it, it’s not a mistake, it’s a choice.”
Steven Denn

Something that has been kicking around my mind for the last little while is the relationship between responsibility and self-compassion. A couple of people recently made some very pointed observations about my lack of self-compassion, and it provoked a strange sadness in me. Sadness because while I know that their point is true – I am often very hard on myself, to the detriment of my happiness – those thought patterns seem so philosophically necessary that I have been unable to change them. This post is my attempt to unpack and understand that philosophical necessity.

Our ability to feel compassion is intimately tied to our judgement of responsibility in a situation. If you get unexpectedly laid off from your job, that’s terrible luck and most people will express compassion. However, if you were an awful employee who showed up late and did your job poorly, then most people aren’t going to be as sympathetic when you finally get fired. As the saying goes: you made your bed, you lie in it. More abstractly, we tend to feel less compassion for someone if we think that they’re responsible for their own misfortune. This all tracks with my lack of self-compassion, as I have also been told that I have an overdeveloped sense of personal responsibility. If I feel responsible for something, I’m not going to be very compassionate towards myself when it goes wrong; I’m going to feel guilt or shame instead.

Of course, this raises the question of what we’re fundamentally responsible for; the question of compassion is less relevant if I actually am responsible for the things that are causing me grief. People largely assume that we’re responsible for our own actions, and this seems like a reasonable place to start. It makes sense, because our own actions are where we have the clearest sense of control. While we can control parts of the outside world, that process is less direct and less exact. Our control over ourselves is typically much clearer, though still not always perfect.

If we assume that we have control over ourselves and our actions, this means that we also have responsibility for ourselves and our actions. If we avoid or ignore that responsibility and we’re not happy with the consequences, we don’t deserve much, if any, compassion: it’s our own damn fault. This all seems… normal and fairly standard, I think, but it’s an important foundation to build on.

Freedom, and Responsibility over Time

Now let’s explore what it means to be responsible for our actions, because that can be quite subtle. Sometimes our choices are limited, or we take an action under duress. Even ignoring those obvious edge cases, our narrative dictates the majority of our day-to-day decisions. What responsibility do we bear in all of these cases? Ultimately I believe in a fairly Sartrean version of freedom, where we have a near-limitless range of possible actions every day, and are responsible for which actions we take. Obviously some things are physically impossible (I can’t walk to Vancouver in a day), but there are a lot of things that make no sense in the current framework of my life that are still theoretically options for me. If nothing else, I could start walking to Vancouver today.

Assuming that we’re responsible for all of our actions in this fairly direct way, we also end up responsible for the consequences of actions not taken on a given day. There is a sense in which I am responsible for not walking to Vancouver today, because I chose to write this essay instead. I am responsible for my decision to write instead of walk, and thus for the consequence of not being on my way to Vancouver. This feels kind of weird and a bit irrelevant, so let’s recast it into a more useful example.

A few hours from now when I finish this essay, I’ll be hungry to the point of lightheadedness because I won’t have eaten since breakfast. Am I responsible for my future hunger? There’s a certain existential perspective in which I’m not, since it’s a biological process that happens whether I will it or not. But it’s equally true that I could have stopped writing several hours earlier, put the essay on hold, and had lunch at a reasonable time. I am definitely responsible for my decision to keep writing instead of eating lunch, and so there is a pretty concrete way in which I am at least partly responsible for my hunger.

This isn’t to say that I’m necessarily going to be unhappy with that decision; even in hindsight I may believe that finishing the essay in one fell swoop was worth a little discomfort. But it does mean that I can’t avoid taking some responsibility for that discomfort. And, since I’m responsible for it, it’s not something I can feel much self-compassion over; if I decide in hindsight that it was a terrible decision, then it was still a decision that I freely made. If I experience a predictable consequence of that decision, it’s my own damn fault.

This conclusion still feels pretty reasonable to me, so let’s take a weirder concrete example, and imagine that in three months I get attacked on the street.

Predictability in Hindsight

I don’t have a lot of experience with physical violence, so if I were to get attacked in my current state then I would likely lose, and be badly hurt, even imagining my attacker does not have a weapon. To what degree am I responsible for this pain? Intuitively, not at all, but again, the attack happens three months from now. I could very well decide to spend the next three months focused on an aggressive fitness and self-defence regimen (let’s assume that this training would be effective, and that I would not get hurt in this case). Today, I made the decision to write this essay instead of embarking on such a regimen; in the moment today, this decision to write is clearly one that I am responsible for. Does this mean I’m responsible for the future pain that I experience? I’d much rather avoid that pain than finish this essay, so maybe I should stop writing and start training!

The flaw in this argument, of course, is that I don’t know that I’m going to get attacked in three months. In fact, it seems like something that’s very unlikely. In choosing not to train today, I can’t accept responsibility for the full cost of that future pain. I should only take responsibility for the very small amount of pain that is left when that future is weighted by the small probability it will actually occur. If I did somehow know with perfect accuracy that I was going to be attacked, then the situation seems somewhat different: I would feel responsible for not preparing appropriately in that case, in the same way I would feel responsible for not preparing for a boxing match that I’d registered for.

All of this seems to work pretty well when looking forward in time. We make predictions about the future, weight them by their probability, and take action based on the result. If an action leads to bad results with high probability and we do it anyway then we are responsible for that, and don’t deserve much sympathy. We rarely go through the explicit predictions and calculations, but this seems to be the general way that our subconscious works.

But what about looking backward in time? Let’s say I decide not to train because I think it is unlikely I will be attacked, and then I get attacked anyway. Was this purely bad luck, or was my prediction wrong? How can we tell the difference?

Depending on the perspective you take you can get pretty different answers. Random street violence is quite rare in my city, so from that perspective my prediction was correct. Hindsight is a funny thing though, because in hindsight I know that I was attacked; in hindsight probabilities tend to collapse to either 0 or 1. And knowing that I was attacked, I can start to look for clues that it was predictable. Maybe I realize that, while street violence is rare in the city overall, I live in a particularly bad neighbourhood. Or maybe I learn that while it’s rare most of the time, it spikes on Tuesdays, the day when I normally go for a stroll. If I’d known these things initially, then I would have predicted a much higher probability of being attacked. Perhaps in that case I would have decided to train, or even take other mitigating steps like moving to a different neighbourhood. Who knows.

What this ultimately means, is that I can only possibly be responsible for being hurt in the attack if I’m somehow responsible for failing to predict that attack. I’m only responsible if for some reason I “should have known better”.

Taking Responsibility for our Predictions

[I should take this opportunity to remind people that this is all hypothetical. I was not attacked. I’m just too lazy to keep filling the language with extra conditional clauses.]

At this point I’ve already diverged slightly from the orthodox position. A large number of people would argue that my attacker is solely responsible for attacking me, and that I should accept no blame in this scenario. This certainly seems true from the perspective of law, and justice. But in this essay I’m focused on outcomes, not on justice; ultimately I experienced significant pain, pain that I could have avoided had I made better predictions and thus taken better actions.

Let’s return to the issue of whether or not I can be held responsible for failing to predict the attack. There is a continuum here. In some scenarios, the danger I was in should have been obvious, for example if my real-estate agent warned me explicitly about the neighbourhood when I moved in. In these scenarios, it seems reasonable to assign some blame to me for making a bad prediction. In other scenarios, there was really no signal, no warning; the attack was truly a stroke of random bad luck and even in hindsight I don’t see a way that I could have done better. In these scenarios, I take no responsibility; my prediction was reasonable, and I would make the same prediction again.

As with most things, practical experience tells me that real-world situations tend to fall somewhere in the middle. Nobody’s hitting you over the head with a warning, but neither is the danger utterly unpredictable; if you look closely enough, there are always signs. It is these scenarios where I think that my intuition fundamentally deviates from the norm. When something bad happens to me, then I default to taking responsibility for it, and I think that’s the correct thing to do.

Of course, there are times when whatever it is is my fault in an uncontroversial way, but I’m not talking about those. I’m talking about things like getting attacked on the street, or being unable to finish a work project on time because somebody unexpectedly quit. These are the kind of things that I expect most people would say are “not my fault”, and I do understand this position. However, I think that denying our responsibility for these failures is bad, because it causes us to stop learning. Every time we wave away a problem as “not our fault” we stop looking for the thing we could have done better. We stop growing our knowledge, our skills, our model of the world. We stagnate.

This sounds really negative, but we can frame it in a much more positive way: that there’s always something to learn, and that we should always try to be better than we were. What I think gets lost for a lot of people is that this not a casual use of “always”. Even in failures that are not directly our fault, there is still something to learn, and we should still use it as an opportunity to grow. Unless we perfectly predicted the failure and did everything we could to avoid or mitigate it, there is still something we could have done better. Denying our responsibility for our bad predictions is abdicating our ability to grow, change, or progress in life. Where does this leave us?

Any good Stoic will tell you that when something goes wrong, the only thing we have control over is our reaction, and this applies as much to how we assign responsibility as to anything else. We are responsible for the fate of our future self, and the only way to discharge that responsibility is to learn, grow, and constantly get better. The world is full of challenges. If we do not strive to meet them, we have no-one to blame but ourselves.


New Comment
8 comments, sorted by Click to highlight new comments since: Today at 1:12 AM
This all tracks with my lack of self-compassion, as I have also been told that I have an overdeveloped sense of personal responsibility. If I feel responsible for something, I’m not going to be very compassionate towards myself when it goes wrong; I’m going to feel guilt or shame instead.

Perhaps you're not for something (like blame), you're against something else (like not improving).

Responsibility: against (not giving it your all), for (figuring out ways to improve)

Things to consider:

The Future: against (blame), for (making changes that increase effectiveness)

If you find blame useful because it motivates/reminds you to do X, then finding some other way to do X without blame (if blame is painful) might make the blame unnecessary.

we tend to feel less compassion for someone if we think that they’re responsible for their own misfortune

This is a matter of perspective. A person may be responsible for doing X, and yet not be responsible for having the traits that predictably push them towards doing X. Everyone is a victim if you go meta enough. Ultimately, feeling compassion is a choice. And so is feeling self-compassion.

[-][anonymous]3y 1
Everyone is a victim if you go meta enough

There is a sense in which this is true, but also a sense in which society doesn't actually function like this. In practice, very few people think meta enough to discard the idea of responsibility entirely. I think my description of how people tend to function is still true (as much as any so general claim can be).

Sure. But how the society actually functions is... well, the usual monkey politics. It is optimized for all kinds of things you might not approve of when described honestly. Such as feeling compassion towards people who might become useful allies in the future. Or feeling compassion towards people who are unlikely to become competitors in the future, so that you can signal some desirable traits about yourself. It is definitely not optimized for providing "fair" feedback.

So, if you choose to follow the same rules for self-compassion... well, it may be good if you try to predict other people's compassion towards you (although, whom exactly are you trying to predict? the answer may depend). But there is nothing inherently "right" about it. For example, I suppose that people who feel high-status will find it much easier to forgive themselves than people who feel low-status (because that is the correct prediction of how likely the others would forgive us). Also, you have the gender bias, where women are bombarded with messages how they need to learn to love themselves, while for men Stoicism is the only approved option. Are you going to internalize this? Again, your choice. But it may be a different choice whether you optimize for "what the society wants me to feel" or "what feeling is best for me".

This is essentially the problem of victim blaming. Obviously it's different since you're using a self-orientated perspective instead of an other-oriented perspective, but the concept of deciding whether blame can be assigned for not having avoided a scenario is the same.

My understanding is that one argument against victim blaming is that we shouldn't expect everyone to constantly be monitoring their situation (always striving to meet challenges) because it isn't possible for humans to do that. We get tired. We miss something. We forget. And then if you're in a mindset where you blame yourself if anything goes wrong, you will inevitably blame yourself for something because it isn't possible to everything right. This blame can easily lead into a spiral where the self-blame causes guilt/shame which causes stress which causes more mistakes which causes self-blame. Because of human psychology, striving to meet every challenge might just be the thing that causes us to fail at them. We aren't responsible for the fate of our future self because the human brain isn't omnipotent over itself.

Of course, you're welcome to try, but be prepared for burn out.

I don't know, I feel like this post is specifically about something that is not victim blaming and seeks carefully to avoid that territory. That is, I see victim blaming as mostly about the social response to blame while this is about trying to assess the causal factors in an outcome. Whether or not one chooses to interpret those causes as blame depends on norms that are a bit besides the point of much of this analysis.

You and I are quite similar, in that we both default to taking responsibility for the painful outcomes that come our way. I do think though that there is a flaw in one little bit of this mindset (or maybe it's just the way you expressed it, so tell me if I didn't get what you were pointing at). It sounds like your formulation is basically "if an outcome is my responsibility, then I refuse to give myself compassion." It is possible to accept a high degree of responsibility while still being compassionate to yourself. I know, because I have seen others do it well, and for my own mental health, one of my goals has been to get better at this.

I always find this much easier to apply to myself when I imagine that I'm applying it to a friend, or to a clone of myself, as I tend to be more compassionate to others than to myself. Could their fate have been avoided if they had been more vigilant? Yes, almost always. But you can't be vigilant about everything, and you still need to take a break from self-flagellation long enough to actually live and get things done.

I might be entirely misunderstanding what you mean by compassion, though. The times that I have been less compassionate to myself than I would be to a stranger have usually been fairly extreme. It has been very valuable for me to do less of that kind of shame spiralling. Unproductive responsibility and shame spiralling can still sometimes be induced by someone important telling me that I am responsible and that I should feel bad. That's a whole different set of problems though.

That doesn't mean that I take any less responsibility, or don't try as earnestly to learn from pain. Making ourselves suffer for our mistakes sometimes feels like a necessary step toward learning, but it does not have to be.

[-][anonymous]3y 1
"if an outcome is my responsibility, then I refuse to give myself compassion."

I wouldn't phrase it in terms quite this absolutist, but I would say that when something is my responsibility I feel like I deserve much less compassion. That's pretty much what I say in the second paragraph: "we tend to feel less compassion for someone if we think that they’re responsible for their own misfortune".

Making ourselves suffer for our mistakes sometimes feels like a necessary step toward learning, but it does not have to be.

This is where my mind has been wandering post writing this essay as well. Is there a way to continue learning and reminding ourselves to improve our vigilance without the self-destructive self-flagellation aspect? There might be, which would be great.

New to LessWrong?