You do not need 'ethics' to be good

Good interactions involve knowledge, skill, care, and proximity

Taddeo di Bartolo, Gluttons in Hell (1393)

As a dislocated fourteen year-old, I desperately wanted to know “how to be good”¹. Luckily, I found refuge in moral philosophy. It taught me that I could learn to be good by thinking through the right rules. These could be found by logically examining thought experiments:

• Should you divert a runaway trolley to kill one person instead of five?

• Is it ethical to steal bread to feed yourself?

• If you and ten people are starving on a boat stranded at sea, is it wrong to use a lottery to figure out who you should eat? 

How wonderful, I thought, if I can figure out these problems, I’ll be more good. 

Besides being an annoying atheist, my favorite pastime as a teenager was reasoning through the moral dilemmas of brain cancers that made you want to smoke, of children drowning in ponds, and of living in a simulation.

I spent more years reading about philosophy, and eventually studied with Professor Caspar Hare, who had taught 24.0x, the virtual class that formally introduced fourteen-year-old me to philosophy.

By the end of my undergraduate degree, after thousands of hours of engaging with moral philosophy, I was still confused:

1. There was no principled means of resolving disagreements

At the heart of the deep moral disagreements in the history of philosophy was a flappy nebulous cloud. So how do you know if it’s actually right to divert the trolley? It would come down to “common moral intuition”, or what’s “reasonable”. It troubled me that what was ‘reasonable’ changed over time. It also troubled me that few philosophers acknowledged the contingent nature of ‘reasonable’.

2. Philosophers did not seem any better than other academics

I noticed that my friends who thought more about philosophy were not more kind, nor were the philosophy professors really much different from any of the other professors I had. The empirical work by Schwitzgebel and Rust on the behavior of ethicists² confirmed my own observations: thinking about moral philosophy doesn’t reliably change behavior. 

3. Knowledge of moral philosophy was rarely useful in my life³

Knowledge of ethics did not, for the most part, help me with my moral dilemmas. Kant had very little to say about whether or how to end a particularly difficult romantic relationship, or whether I was being a good friend.

One day I found myself in front of a spreadsheet, trying to rationally assess whether my girlfriend and I should break up. I made up attributes like “physical chemistry” and “social life”, then weighted and multiplied them for a final score. There were many obvious problems as I tried this: although I could make up numbers to get to a score, the score wasn’t stable across time, and neither were my categories. Did I really care about “social life” as a discrete category? Was it more or less important than our intellectual connection? 

In general, I noticed that this type of thinking about what to do took me away from the immediacy of everyday kindness. The kindest people seemed to have goodness spout out of them effortlessly, spontaneously. They weren’t using some logical system

I stepped away from engaging with academic ethics for a couple years. Gradually, I began to feel that I was, at long last, making some progress on understanding what ‘being good; means. 

Mistakes of ethics

Thought experiments are lousy practice for being good

Imagine that you tell your parents you really want to ride a bike. They nod and tell you that it’s an important skill we all need and how it brings autonomy to your life. Then instead of giving you a bike, they give you… books. Books about the history of the bicycle, the physics of how gear ratio affects your pedaling, manuals about all the different types of bikes. 

After studying for ten years, how equipped would you be to ride your bicycle?

I think moral philosophy is the same way. There is nothing in the way of developing procedural skills for being good. Instead what you practice in philosophy class is how to argue. You graduate with a philosophy degree with a bunch of propositional knowledge and the skills for making anything sound logical.

Being good is not (just) knowing facts and applying rules

Knowledge is not just a set of propositional facts. There is a big difference between knowing the capital of France, knowing how to speak French, and knowing what Paris feels like⁴.

The big mistake that most moral philosophers made was assuming that morality came out of deliberation over propositional facts. Propositional knowledge can certainly be part of being good, but alone it is insufficient.

Being good

One of my bigger realizations was that I should stop trying to find or build a complete ethical system. I suspect such a thing is impossible and fortunately isn’t all that important for what I’m calling ‘good’. Instead, I’m more interested in practical heuristics that can bring about more good in the world.  The points below can and hopefully will continue to be refined. Here’s a beginning:

“Being good” is not a property of a person — it’s more helpful to think about “good interactions”.

Rather than focusing whether a person is good or bad, or even ‘the right action to take’, I like to think about ‘cultivating the environment for good interactions’. Like a gardener providing the ingredients for seeds to sprout, we can shift the emphasis from the exact rules governing an action to the factors involved in a situation. 

‘Good’ arises out of a definite context

Most ethicists have a very abstract notion of ‘ethical’. Bentham’s fixation on maximizing utility is one good example. Instead of thinking of good as some quantity or quality in an abstract realm⁵, I want to emphasize that it arises for everyone in their specific context. Good arises wherever you are right now.

It’s closer to the spontaneous offering from your roommate after he spent three hours making risotto, the care in listening deeply to your grandmother recount her childhood, and even the pull request you submit to the maintainers of your favorite open source software package. 

“Good interactions” involve contextual knowledge, care, skill, and causal proximity. 

Care — for good interactions to arise, it's important that there's warmth in how you engage with the other. Feeling love, deep care, or compassion towards the stranger or your friend will make a big difference. This is something that can be cultivated⁶.

Contextual knowledge — we need to understand the details of the specific situation to help. Every year, hundreds of dogs are sent to an emergency vet because of chocolate poisoning. Presumably, most of the owners intended to make their dog happy, rather than poison it. Because they didn’t have the contextual knowledge that dogs, unlike humans, cannot metabolize theobromine, they end up harming their dog. Similar dynamics apply to humans helping other far away humans.

Skill — we also need skills to be able to bring about good. 

“I’m so sorry about how your grandma died — drowning in her lung fluids over five days — God that must have been so painful and horrible”. The hypothetical person speaking this meant well, but, lacking the skill of attuning to the situation, spoke in a way that evoked discomfort. It takes skill for intentions to translate to successful interactions.

Causal proximity — you need to actually be able to interact with the situation. 

This is a basic precursor but is worth mentioning — you’re not going to be able to help blind children on the outskirts of Jakarta if you have never left Ohio if only for the simple fact that your actions cannot affect them.

So how is this understanding of ‘good’ practical? 

It directly informs what it means to cultivate more good in the world. Rather than thinking and reading abstract arguments about systems, this breakdown implies that you would benefit from going out and being in the world with an eye to increasing skill, knowledge, and care. 

Were I allowed to send a brief paragraph to fourteen-year-old Max, I’d say something like this:

Moral philosophers like J.S. Mill are more helpful for understanding human culture than for how to bring about the good. This is important contextual knowledge but no substitute for care or skill. So, in addition to reading the philosophers, travel broadly and work closely with people. Whether you end up in Islamabad or West Virginia, become locally involved. Understand what it feels like to be grow up in a community neglected by the rest of the country, figure out how to start a business. The nature of the good has to come through in your local situation, so get embedded and prioritize solving problems for people in your life over abstract thought experiments.

P.S. I still donate to GiveWell. I don’t entirely reject that style of consequentialist thinking. There’s at least another essay in making what I’ve written about here compatible with this, which I may or may not get to.

P.P.S. I experimented with recording an audio version for the last couple of essays. If anyone thinks this is helpful, let me know and I will continue.

Thanks to Tuomas Oikarinen, Nick Stares, and Jake Orthwein for their companionship in our journey into the mess of ethics.

1.  I suspect it has something to do with me being a painfully self-aware foreigner in a new country with an out-of-place accent.

2.  “We review and present a new meta‐analysis of research suggesting that ethicists in the United States appear to behave no morally better overall than do non‐ethicist professors. Measures include: returning library books, peer evaluation of overall moral behavior, voting participation, courteous and discourteous behavior at conferences, replying to student emails, paying conference registration fees and disciplinary society dues, staying in touch with one's mother, charitable giving, organ and blood donation, vegetarianism, and honesty in responding to survey questions.” Schwitzgebel, Eric, and Joshua Rust. "The behavior of ethicists." A companion to experimental philosophy (2016): 225-233.

3. Here I want to add one significant exception. After encountering the arguments of Peter Singer, Toby Ord, and Will MacAskill, I was swayed by an undergraduate into thinking deeply about how to mitigate risks from new technology (at first AI, then biosecurity). This led me into AI research early on, then my work in technology policy. My relationship with this style of thinking has changed significantly since then, as I gesture at in this essay. I may write at more length about this.

4. We can separate types of knowledge into propositions, procedures, and perspectives, to borrow from John Vervaeke in his "Relevance, meaning and the cognitive science of wisdom". He in turn, was building on John Dewey in his “Human Nature and Conduct”.

5. Part of the problem of Parfit’s Repugnant Conclusion is that in abstracting ‘happiness’ from a concrete context, he fails to appreciate its embedded nature. Numbers are a tool we’ve used to great success in much of science and engineering, but applying them to an abstract thought experiment fails to appreciate the good.

6. Read widely, make many friends, fall in and out of love — care is deeply in our nature as human beings. Contemplative compassion practices are also a fantastic supplement, if you find some that are a good fit