The claim can be quite hard to argue against because it has a certain vagueness that makes it hard to pin down what (if anything non-trivial) is really being claimed. Regardless, here are eight arguments I put together against the idea that everything we do is “selfish” or just for our own pleasure:

1. There are many actions that we take automatically and thoughtlessly due to repetition – not because of wanting or pleasure. Consider habits like brushing our teeth or sticking our phone in (for example) our left pocket (as opposed to our right one) when we’re done using it . These sorts of behaviors can be so automatic that we forget immediately afterward whether or not we’ve done these things. The point is not that these habits aren’t useful, just that at some point, we come to do them automatically without even considering whether they benefit us or not (as you may dramatically learn as you drop your phone on the ground when wearing shorts without pockets, thoughtlessly executing the “phone in pocket” habit).

Some other examples of automatic behaviors that we usually do without having any apparent desire/wanting/pleasure/pain involved: breathing, continuing to walk (once we’ve begun to walk), balancing, swallowing (once food is chewed), looking towards a sound (when there is an unusual but non-threatening noise), and social mirroring of body language.

2. Evolution didn’t select for humans based on how much they did what they “wanted” or based on who maximized their own pleasure. Rather, it selected for those whose genes spread most. Selfishness and pleasure are important tools that evolution used, but not the ONLY motivator. For instance, genuine altruism and a sense of obligation towards kin and allies can provide substantial evolutionary advantages!

3. Which pleasure are we talking about? For instance, we clearly sometimes forgo more pleasure now in exchange for extra pleasure later (e.g., by getting work done early) and other times sacrifice long-term pleasure for the short term (e.g., playing video games instead of studying when you have a big test tomorrow). So if people are just maximizing for their own pleasure, which pleasure are they maximizing for?

You might be tempted to reply, “that’s because we’re just adding up the pleasure across time to decide what to do.” But it seems clear that some people aren’t doing this (e.g., drug addicts who know their life is being ruined but sacrifice everything for the next fix). Many experiments in behavioral science and behavioral economics also contradict the idea that people are merely happiness maximizers. It’s too simple to say we “just do what gives us pleasure.”

4. “Wanting” shouldn’t be conflated with something bringing pleasure or reducing pain. They are quite correlated (since we tend to want pleasurable things), but pleasure and wanting are distinct. There are things we can really want (e.g., to “one day understand a mysterious scientific principle,” or “to keep promises” or to have certain things happen after we die) which are not about our pleasure.

Some neuroscience papers claim that “wanting” and “liking” can even be separately stimulated in rat brains. Whether scientists know how to do this or not, it seems we sometimes want things because they bring pleasure, but other times we just WANT them, so wanting and pleasure are not identical.

5. If the claim is weakened to say that we humans always do things that we have SOME sort of motivation to do, then (interpreting “motivation” broadly) the claim is trivially true. But it also doesn’t say anything – it’s right just by definition. Motivation is not identical to pleasure or wanting. So defining “wanting” to do something as having ANY motivation to do a thing doesn’t work because it renders the argument trivial. Similarly, if “self-interest” or “wanting” is just defined to be any pattern of brain activity that causes us to act, or any form of motivation at all regardless of what sort it is, then it is true (by definition) but also adds no information. What’s the point of even making the claim if it’s true by definition? In those such cases the claim can be actively misleading because “self-interest” has connotations to most people (even if you try to define those connotations away).

6. People sometimes do things that they know will bring them more pain than pleasure. For instance, a protestor who uses gasoline to set himself on fire might feel a spark of pleasure just before he lights the match, but he knows he will tremendously suffer until death. Or consider someone who takes an action for a social cause even though they know it will likely lead to spending the rest of their life in prison. Clearly, the person is sacrificing more happiness than they are gaining by such an action, yet some people do act in this way.

7. If we imagine a person who is extremely altruistic because they love making others happy, and we claim they’re “selfish” because they are doing it just to feel good, this is a very non-standard way to use the word “selfish.” It insinuates their behavior is somehow less good and is misleading in conversation. Of course, we can define words however we want, but if we define them in a way that is different than how others use a word, it makes discussion difficult and confusing.

What work does the word “selfish” do to explain things here? It’s clearer to just say (in this case) “the person is motivated by their love of helping others” and leave it at that. Most people would call that “altruism” (not “selfishness”) upon knowing all the details.

8. When we try to make the claim precise, it’s hard to do so (and, unfortunately, few I’ve encountered making the claim bother to try). We’re clearly not always maximizing long-term pleasure, but nor are we always maximizing immediate pleasure. Claiming we “always do what we want” is not the same as claiming “we always do what is pleasurable.” So maybe we just try the former claim?

If we try to restrict the claim to not be about pleasure or pain and just say, “we do what we want,” then how do we explain our numerous subconscious behaviors? And how do we define “want”? People often say they didn’t do what they “wanted,” so we can’t use colloquial definitions.

If “want” is broadened too much, then we’re back to just claiming that we do what we’re motivated to do; that is, we’re making a trivial definitional claim. So what is really being claimed by “we only ever do what we want”?
I think either nothing interesting or something false.

Now, all of this being said, clearly, people often DO act based on however they “want” to act (by a reasonable definition of “want”). And very often, people do act in such a way as to seek pleasure and avoid pain. It’s just that not all human actions fit that description, which is what the “everything is selfish” crowd claims.

To finish up, I’ll attempt to take the other side/steel-man the claim that “everything we do is something we want.” I think there is a psychological state of “desire for things to be a certain way” that drives many (though not all) of our actions. This desire for things to be a certain way is not the same as pleasure (though is often related to it) and the way we want the world to be is not always the way we think will make us happiest (though it often is). So, although I think that the generalization made by psychological egoism is false, I do think it’s approximately true, in a certain sense, a decent amount of the time.

This essay was first written on November 9, 2021, and first appeared on this site on November 12, 2021.

This way of deciding between options is quite common. When someone asks us whether we want an apple or an orange, we often will imagine what the experience of each is like, and see which simulated experience feels the most positive. So sometimes when we say that we “want” something, what we mean is that when we imagine that thing it produces a positive experience.

Perhaps it was no coincidence that my memory of swimming involved the peak intensity of the experience (jumping into the cold water was a more intense experience for me than any other aspect of swimming). There is a rule of thumb known as the peak-end rule which says that we base our judgement of how good or bad experiences are mostly on their peak intensity and the way that they end. So if an experience is quite pleasant on average but contains a few seconds of very negative experience and ends poorly, we are likely to view that experience after the fact as having been unpleasant. For swimming, not only is the peak intensity rather unpleasant for me, but the ending is as well. It often involves walking around wet, cold, and smelling of chlorine.

Some remarkable studies have been performed that confirm half of this peak-end rule. They involve giving 50% of a group of people (selected at random) an unpleasant experience, and the other half the same unpleasant experience with a shorter, somewhat less unpleasant experience tacked on the end of the primary experience. The latter group views the entire experience as having been less unpleasant than the former group does, and would be more willing to subject themselves to the experience again. This holds even though the latter group had more total moment to moment unpleasantness than the former group, and were subjected to unpleasantness for longer.

A similar issue to my swimming one can arise for experiences that we have never actually had before. When deciding whether we want them, we imagine what they are like, but this can be quite different from what they are in fact like. I have heard it said that this arises particularly often for sexual fantasies. People believe that they would really enjoy a particular experience (they do, in fact, enjoy imagining having that experience) but when they actually have it they find that it is unenjoyable in practice. And something akin to this is at risk of happening when people consider the question of whether they want to have children. If the visual image that the thought of having kids produces is one of sitting by a fireplace while beaming proudly at two well-behaved children with beaming smiles, then the feeling produced may well be “I want that!” But if the image is instead one of changing diapers, or trying to get your kids to stop screaming, or being woken up by a baby after two hours of sleep, the resulting feeling of wanting could be quite different.

Consider the problem that arises from these considerations. When we decide whether we want something, we often rely on our simulation or memory of that thing to make a decision. But our simulations and memories of experiences are not the same as the experiences themselves, and can be inaccurate representations. When we imagine something we run the risk of not understanding what that experience is really like. Therefore, when we say that we want something, we may only want it because of the particular aspects of it we happened to simulation, or because of inaccuracies in our simulation. However, if we force ourselves to think of many different aspects of the experience in question, rather than just one small part, that may lead to less arbitrary conclusions about what it is that we want. If your wanting depends heavily on the fact that (by chance, or due to a certain design feature of the human brain) you happened to imagine a particular part of the experience rather than another part of it, then the fact that you think you want it may not be very meaningful.

Nowadays, when I think about whether I want to swim, I try to get myself to imagine what the vast majority of the experience is like, not just the beginning and the end. I imagine actually swimming in the pool, which produces a positive mental state, rather than only thinking about jumping into the pool, which produces a negative one. As you’d expect, this produces a stronger feeling of wanting to swim than I had before.