Feynman's Greatest Insight

What America's Greatest Scientist Can Teach Us About Learning and Learning to Teach

When it comes to learning, Richard Feynman kinda knew what was up. He won a Nobel Prize for his work in Quantum Electrodynamics. So, he definitely knew some stuff.

But… a lot of people know a lot of stuff. A lot of people have won Nobel prizes.¹ According to the current physics ranking, he’s not even in the top 5 physicists of all time.² Those things aren’t what made Feynman one of a kind. Here’s his most outstanding attribute: Feynman is probably the best science teacher of all time.

His published lectures (“The Feynman Lectures on Physics”) is considered one of the best popular science publications of all time, and he has been given the nickname “the great explainer.” The method of diagramming subatomic particles that he invented has gone global, and a bunch of his teaching illustrations are still used in physics classrooms all over the world.

I wrote in an earlier post that learning how to learn is the most powerful skill a human can develop. And this is why I’m interested in Feynman. He developed a strategy for learning anything that is a really powerful and practical tool for those of us interested in learning how to learn.

As a side note—and this is what really won me over about his theory—his method taps into the deep and spooky recursiveness of learning/teaching that I mentioned in that first post. His theory is summarized by an aphorism he loved: “once taught; twice learned.” The point is that if you really want to know something, try teaching it.

The method itself is really interesting and has become quite popular (try googling it), so I won’t spend too much time recounting it in detail. For now, I’ll just sketch his method,³ and spend more space reflecting on why I think it’s so powerful.

The Feynman Method

1. Write down what you know and what you don’t know about the topic.

2. Teach it to someone else at a 6th-grade level of understanding.

3. Take note of gaps, misunderstandings, and deficiencies in your model.

4. Return to source materials/data to fill in gaps and refine your understanding.

5. Further simplify, refine, and repeat.

Conclusions

Teaching facilitates learning for at least 3 +1 reasons.

First, the ability to simplify complex ideas is one of the best ways to test for deep understanding. Can you explain multiplication to a 6th grader? What about Christology? If we truly understand something, we should be able to articulate its basic architecture in very simple terms, even if that shaves off some nuance.

The process of simplification forces deep thinking about fundamental problems. It prevents us from hiding our misunderstandings behind technical language or appeals to mystery.⁴ Feynman himself differentiated between teaching ideas and merely teaching definitions, and I think that gets right to the point I’m trying to make.⁵

Of course, complex ideas resist simplification beyond a certain level, but that level is usually way simpler than we think. Often, the barrier to simplification is an overly complex model. And the point is that—sometimes—are models are so complex because we don’t understand. If we could just get to the point, we wouldn’t have to ramble on and construct definitions.

A very brief illustration: in a recent interview with Scientific American, nobel prize winning physicist Gerard ’t Hooft suggested that the reason we haven’t made much progress in understanding quantum mechanics is that the model is flawed:

“I’d argue that superpositions of states are not real… things never superimpose. [Erwin] Schrödinger asked the right questions here: You know, take my cat, it can be dead; it can be alive. Can it be in a superposition? That’s nonsense!”

Some of the greatest breakthroughs in physics, for example, have been simple models—new conceptual frameworks—that cut through seemingly intractable problems. You gotta be willing to reset. These moments in the history of math or physics tend to be catalysts for huge developments because they restart the conversation from a new, clearer, conceptual starting point.

Now, those models are still more complex than the average 6th grader could understand, but the point is that simplification—striving for it, seeking it, valuing it, rather than getting caught up in quarrels about words—actually unlocks and facilitates deep understanding. Even if you aren’t re-writing the sciences, avoiding technical excess in your own thinking and help you get to the essentials.

Second, when you try to externalize what you know, you realize what you don’t. If you’re like me—good at pretending to know things—you tend to downplay what you don’t know. Obviously, that’s not great for a few reasons… but the relevant point here is that deception runs deeper than mere appearances. We trick ourselves into thinking we know more than we do.

It’s easy to read a new idea and assume you’ve understood. That is much more comfortable than really being honest about the ways that a new idea eludes or confuses you. So, we often paper over real deficits in understanding without even truly realizing it.

But teaching exposes those deficits. Externalizing our knowledge is a way of shining light on our self-deceptions. When you try to explain it out loud, you realize you can’t. Or you realize you can’t without using big words. As a bonus, studies show that this process of recalling and putting ideas into your own words is a powerful way to transmit ideas into long-term memory.

Third, and relatedly, having someone else ask questions is another way of getting around our epistemic blind spots. Even when we’ve sufficiently externalized an idea, given a gap-free explanation that satisfies us, someone else might see a hole we hadn’t. That means that presenting work for feedback and questioning supercharges our learning.

If it’s not already obvious, these are some of the main reasons I started a substack.

Last, the recursive nature of his method is so wise. It’s wash, rinse, repeat; not repeat until “smart.” It’s especially striking because Feynman was an expert among experts. He’s the SGOAT.⁶ And yet, his method doesn’t boil down to, “do this until you are an expert like me.” Instead, it’s more like, “this is how you get started in a life-long process.” And I love that. Perhaps the main reason Feynman was such an excellent teacher was that he never stopped being a learner.

1. Not really that many tho.

2. He sits at an embarrassing 7th place. https://www.caltech.edu/about/news/physics-world-poll-names-richard-feynman-one-10-greatest-physicists-all-time-368

3. Apparently, there’s no record of him actually teaching this as a method of learning, although lots of his quotes and anecdotes about him seem to indicate that he thought of learning this way. More here: https://hsm.stackexchange.com/questions/16096/feynman-learning-technique/16097#16097

4. This was a shot at lazy theologians.

5. This finally gave me language for why I’m allergic to “Christianese”

6. Seventh greatest of all time. Understanding the acronym is a reward for reading the *endnotes* (it annoys me that substack calls them footnotes when you have to scroll to the bottom to read them).