by Brian Tomasik
First published: 9 Feb. 2017; last update: 23 Oct. 2017
This piece reflects on my personal experiences with different sources of altruistic motivation. "Pain-based altruism" comes from horror at the suffering of others and a parenting-like impulse to relieve that suffering as quickly as possible. "Enjoyment-based altruism" is motivation toward altruistic ends by factors that aren't intrinsically altruistic, such as the fun of learning or the desire for social standing. I feel that both of these motivations are needed in their own ways.
When I was in high school, one of my teachers announced a test on the material we had been reading. The class responded with a collective groan and asked why we needed a test. The teacher replied with a statement to the effect that "Fear is a stronger motivator."
There's abundant literature on motivation by carrots vs. sticks, most of which I haven't read. This piece discusses some of my current introspections about the roles of guilt vs. fun-seeking in motivating my own altruism. What I say in this piece has probably been thought of many times by many authors.
The importance of empathetic pain
Some effective altruists report that their work is motivated by excitement about an opportunity to help others.
- Holden Karnofsky: "I expect a growing number of people to be asking the question, 'How can I make the most of this opportunity?' And I hope they’ll ask it not from a place of guilt and obligation, but from a place of self-actualization and excitement."
- Leah Libresco: "I think the thing is that you have that sense of a thrilling opportunity when it comes to effective altruism and to distance charity, that — because your dollars go so much farther abroad — you’re being invited to do an enormous amount of good, and that’s exciting, like, it’s not just like this grim obligation, it’s exciting!"
While I share these sentiments to some degree, my own experience is that altruistic motivation very often starts with sadness and guilt: i.e., "things are really bad for these other organisms, and how can I continue to focus on selfish interests while they're suffering?" Without feeling this kind of altruistic pain, it's not clear to me that I would have enough motivation to focus on one exciting opportunity (altruism) rather than numerous other equally exciting and spiritually rewarding projects. Suffering is a call of urgency that trumps other concerns. (Incidentally, this is also why I'm roughly a negative utilitarian rather than a traditional utilitarian. Nothing else carries the same moral weight as intense suffering.)
Guilt-based altruism is like attending to your baby when it's crying. You don't do it because it's "exciting to have the opportunity to comfort your baby". You do it because while your baby is distressed, nothing else matters until you relieve the baby's suffering.
Seeing widespread suffering in the world -- from political prisoners tortured in China, to ladybird beetles overwintering in my house whom I accidentally half-crush under my shoe -- is like watching one's baby crying. One's immediate reaction is to feel distressed and less interested in trivialities while the problem is ongoing. Contributing to relief of suffering is a way to relieve oneself of the vicarious pain that compassion creates. This is why altruism is so viscerally important to me. It's not about generously helping others "out there". It's about reducing the spiritual pain that I myself feel when thinking about all the horrors that the world contains.
Why not ignore all the bad things in the world? Isn't ignorance bliss? This is where guilt comes in. If I tried to permanently ignore the suffering of others, I would feel guilty about doing so. Guilt plays the role of an anti-wireheading mechanism, preventing me from burying my head in the sand and focusing only on the relatively suffering-free bubble of my immediate surroundings. I find that guilt also helps my epistemic rationality more generally, since if I were to hold inaccurate but comforting beliefs, I would be hurting others (in expectation and in the long run) by doing so.
It also helps that even my own house and yard are far from suffering-free, since they contain numerous bugs in various forms of distress, such as flies being eaten by spiders or dying of exhaustion after buzzing at the windows. When I used to walk to class or to work on rainy days, I would see tons of earthworms that had come to the surface and were trampled by inattentive pedestrians. This caused me a regular dose of anguish that contributed to my visceral determination to reduce invertebrate suffering. Without feeling that vicarious pain on a regular basis, I would probably be less attached to the agony of invertebrates.
The downsides of empathetic pain
Seeing your baby in distress causes you to put down other tasks and immediately seek to help. This is a powerful form of motivation, but it also leads to short-term-focused kinds of helping. A crying baby can usually be succored right away. Long-term investments in the baby's health, such as from gathering food or building shelter, can wait until the immediate crisis has subsided.
Unfortunately, in the world at large, the crisis never subsides. There is always far more intense agony going on than we can even imagine, much less avert. This may lead us to feel always stressed, causing compassion fatigue and an inability to relax.
The mindset of "doing something to help now!" also favors short-term actions driven by emotional need rather than more far-sighted investments. For example, whenever I'm confronted by the tragedy of worms being crushed on the sidewalks, I demand of myself that I take some time to help move worms out of harm's way, because it would be heartless to just leave these worms right in front of me to get squished. Rescuing individual worms in this way is unlikely to be the most cost-effective method to relieve invertebrate suffering, but it feels more urgent than reading obscure facts about earthworm physiology or writing comments on someone's Facebook post. Longer-term actions to help invertebrates, such as building a movement of people concerned about wild-invertebrate suffering and pushing for environmental policies that reduce the number of invertebrates born, require many slow steps without immediate payoff, making them feel somewhat trivial in the face of an urgent crisis.
Finally, urgent helping impulses can sometimes hinder epistemic and moral open-mindedness. When you're driven to, e.g., spare as many farm animals as you can, you're less likely to be receptive to doubts, such as whether promoting veg*ism might harm wild animals. Anything that gets in the way of your goal may seem like an obstacle; this is what Julia Galef calls a "soldier mindset" rather than a more curious "scout mindset". For the same reason, you might be less worried about intellectual integrity and acknowledging legitimate contrary views. In crisis mode, the ends justify the means, and there's no time to evaluate whether these are actually the correct ends.
An alternative to the picture painted above is to find motivation in fun projects that are also altruistically useful. In my case, because I find learning so enjoyable, this means that I focus on researching topics that are important for reducing suffering. For other people, having fun while doing good may involve movement-building, earning to give, and so on.
With this approach, day-to-day motivation is relatively divorced from the horrors that one is aiming to reduce. Instead, you're driven primarily by ordinary stimuli, like the joy of learning, influence by peers, drives for status, and so on. While you happen to be working toward an altruistic goal, daily life feels pretty ordinary.
I think this is the rhythm that most altruists tend to fall into most of the time in the long run. That's probably good, because this way of living is less liable to cause burnout. It also avoids many of the other pitfalls of pain-driven altruism, such as neglect of long-term investments.
However, there are downsides to enjoyment-based altruism as well. In my experience, the biggest problem is that it's easy to drift into suboptimal activities that seem fun but may not actually be extremely valuable. In my case, this takes the form of spending too much time learning about topics that aren't relevant enough to achieving practical results.
This was less of a problem when I was in school, since in that case, learning about interesting abstract topics is also fairly important for the instrumental reason of getting good grades. Likewise, if you're earning to give, there's a relatively well defined path to impact (namely, doing well at your job). That said, even in these cases, there's uncertainty over how much marginal effort to put into earning good grades or good bonuses vs. how much to put into free-time activism.
Karnofsky said regarding some critics of effective altruism: "I think such people fundamentally misunderstand effective altruism. I think they imagine that we have passions for particular causes, and are trying to submerge our passions in the service of rationality. That isn’t the case." In my own case, the critics are somewhat right. I find some topics particularly interesting, even if they're not the most altruistically important. For example, when I was in high school and college, I was obsessed with math and wanted to do as much of it as I could. More recently, I'm often more obsessed with various other, more qualitative topics. But either way, there are differences in what topics I enjoy most, and those differences don't perfectly align with what has highest impact. This point is even more stark in the case of career choice. I enjoy some careers more than others, and those preferences may not align with what has highest impact. The big problem with enjoyment-based altruism, for me at least, is that it's very easy to slide into just doing what's most fun, and it's often difficult to disentangle what kinds of intellectual and personal investments are actually important vs. which ones are more selfish.
I and others worry that some far-future-focused altruism may fall into the category of intellectual speculation without clear practical application. On the flip side, many far-future-focused altruists validly allege that short-term-focused altruism is driven too much by the need to reduce empathetic pain and not enough by a more dispassionate weighing of expected values in the face of massive uncertainty. Moreover, it's often hard to anticipate how useful knowledge that you don't yet know will be. For example, I now care to a nontrivial degree about bacteria suffering, but if I hadn't taken the time to learn open-endedly about biology and neuroscience, my values may have ossified, causing me to only care about organisms at the complexity of insects or higher. And of course, if my values had ossified even sooner, I might not have cared about insects either.
The tension between pain-based and enjoyment-based motivation is not easily resolved, and it represents a recurring theme in my life. The approach that I've tended toward -- and perhaps the approach that many people tend toward -- is what we might call "two-level altruism", in analogy with two-level utilitarianism. Most of the time, we use enjoyment-based altruism: We follow the routines of daily life, being motivated by interestingness, social incentives, and so on. But every once in a while, we get shocked out of complacency by the horror of suffering. This forces us to assess whether our enjoyment-based pursuits are on the right track and, if not, what high-level steps we can take to realign them (such as changing one's intellectual focus, donation targets, career, social circles, etc.). This general cycle between (1) execution and (2) evaluation + planning is pretty standard, such as in the business world.
I go through one of these (figurative) "midlife crisis" re-evaluation periods about once every few months -- not by design but just because of the random walk that my psychology follows. For a short period of time, I begin to doubt whether what I'm doing is actually very useful and try to open myself up to those unpleasant arguments against my current approach that I normally don't worry too much about. (This process of deliberately considering unpleasant thoughts is partly inspired by Adriano Mannino.) During this compassion-focused phase, I often lose some interest in my longer-term learning, because it doesn't look as useful at reducing suffering right now. This may help to (slightly) weaken my emotional attachments to fun topics that aren't necessarily altruistically optimal.
Unfortunately, because I have limited willpower, the result of my re-evaluation phase is often to roughly confirm the status quo -- not necessarily because I think my current trajectory is optimal but because I think my current trajectory is somewhat close to optimal within the constraints of my (rather significant) selfishness. For instance, part of my brain intrinsically values learning enough that it may selfishly insist on more time spent learning than is altruistically warranted. That said, I sometimes do make subtle shifts in my focus due to these reorientation periods.
Keeping alive both compassion and curiosity
One of the reasons I rescue worms from being crushed by pedestrians, even if doing so isn't necessarily optimal in terms of direct impact, is that I worry about suppressing my empathetic tendencies. I worry that if I cut off the anchor of altruistic pain, I might end up without any way to prevent my ship from drifting on the ocean, pushed by the winds of enjoyment-seeking. Libresco expresses a similar point:
I don’t actually want to have the practice of saying no to the friends around me all the time; I don’t want to squelch that impulse to love and to care for others. What I want is I want to harness it so that it pulls me onward to people that never make that personal demand of me, whose faces I don’t see; so that this one particular act of love, I’m like, “Oh, right! And if I knew these people, I’d love them too — so let this remind me to act further.”
At the same time, I'm also cautious to avoid quashing my curiosity. When I'm in a pain-focused mode, learning about things other than how to directly make an impact feels trivial. Fortunately, after that phase subsides, I return to a psychological state where I'm open to exploration without explicitly worrying about the bugs I should be saving with these minutes of my time. I even allow a bit of time learning about things that seem on their face to be almost completely altruistically useless, because I want to explore the world and am worried about what might happen if I were to permanently suppress my curiosity itch.
In theory, it would be nice if I could test my policies of not suppressing compassion and curiosity, to see how essential these impulses actually are. Unfortunately, it's risky to experiment on myself for fear of causing permanent damage. So, like superstitious people who perform rituals because they don't want to find out what would happen if they didn't, I don't try fiddling with my compassion or curiosity parameters too much.
Painful empathy vs. warm-fuzzy compassion
Bloom (2014) distinguishes between "emotional empathy—feeling another’s pain" (which "leads to what psychologists call empathetic distress") and "non-empathetic compassion—a more distanced love and kindness and concern for others." As I understand it, this means that rather than being motivated by feeling the pain of others, we can be motivated by the warm fuzzies of knowing that we're helping them.
This is a reasonable suggestion, and maybe most of the time it's better to have the warm-fuzzy form of compassion. However, at least relative to my own psychology, some dose of empathetic distress is important for kicking me out of a self-centered mindset. I can get emotional fulfillment from a number of sources besides helping others, but only helping others can address the spiritual anguish that I feel when mentally simulating others' pain.
Bloom (2014) says:
It is conceivable, I suppose, that someone who hears about the plight of starving children might actually go through the empathetic exercise of imagining what it is like to starve to death. But this empathetic distress surely isn’t necessary for charitable giving. A compassionate person might value others’ lives in the abstract, and, recognizing the misery caused by starvation, be motivated to act accordingly.
Fair enough, but for me, the process of imagining starvation is the strongest source of motivation for caring about starving children. Contemplating that emotional pain is what shocks me out of complacency and makes the issue of child hunger more than another moral abstraction that sounds bad but isn't obviously overridingly important.
Another concern I have with warm-fuzzy compassion is that it may not favor antinatalist methods of suffering reduction. Bloom (2014) explains regarding research by Tania Singer:
ongoing experiments led by Singer and her colleagues in which people are either given empathy training, which focuses on the capacity to experience the suffering of others, or compassion training, in which subjects are trained to respond to suffering with feelings of warmth and care. According to Singer’s results, among test subjects who underwent empathy training, “negative affect was increased in response to both people in distress and even to people in everyday life situations. . . . these findings underline the belief that engaging in empathic resonance is a highly aversive experience and, as such, can be a risk factor for burnout.” Compassion training—which doesn’t involve empathetic arousal to the perceived distress of others—was more effective, leading to both increased positive emotions and increased altruism.
Hoffman (2013), describing a study coauthored by Singer: "Before the training, participants showed activity in an 'empathic' network associated with pain perception and unpleasantness; after the training, activity shifted to a 'compassionate' network that has been associated with love and affiliation."
My concern is that if compassionate people avoid extremely negative feelings, they may be less likely to conclude that some lives are net negative in welfare and should not be brought into existence. Rather, I speculate, people with positive feelings about helping others may be more likely to focus on caring for whatever people or animals end up existing. But preventing the creation of new suffering beings may often be more cost-effective than caring for existing beings, especially in realms like reducing wild-animal suffering and averting far-future suffering. I haven't read Singer's research and don't know if my concern is valid, but it seems plausible a priori.