You Know Much Less Than You Think

The illusion of explanatory depth

3 min readFeb 5, 2021

Of course, you know how a bicycle works, right? Wrong.

Leonid Rozenblit and Frank Kiel, Yale psychologists, first concluded through a series of experiments that people tend to overrate their understanding of causally complex systems like artifacts (bicycles, microwaves, sewing machines, microchips, zippers, refrigerators), and natural phenomena (tides, rainbows, waves, etc.).

They coined a term for this: the illusion of explanatory depth (IOED). Simply put, if you ask a bunch of people to rate their understanding of a causally complex system, say, a bicycle, they are likely to give themselves high ratings. Ask them to explain in detail, with diagrams, how the bicycle works, and they are likely to fumble and eventually rate their understanding lower.

Rebecca Lawson, from the University of Liverpool discovered exactly this, in her study where she asked different groups of people to draw a bicycle.

I think I know less than I thought. — Participant’s comment, after completion of the bicycle drawing task.

Participants’s drawings of the bicycle were often just wrong.

We may understand something conceptually, and yet, fail to understand it in enough detail to be able to explain it.

I have experienced this many times in my own life. When I started teaching Statistics 2, the most basic statistics course at UC Berkeley, I was quite confident about my understanding of stats and probability. Yet, when I started teaching, I found it inordinately difficult to explain the basics of statistics. This experience filled me with shame. I had majored in computer science and math in college. I thought I could simply breeze through my teaching assignment. Turns out that I was experiencing the illusion of explanatory depth all along!

Techniques to fix this illusion

1. Err on the Side of Curiosity

Tim Harford, the writer and host of the podcast, Cautionary Tales, suggests curiosity as an antidote to this illusion. Being more curious could help us gain explanatory understanding of a subject. He also suggests that political debates between right-wing and left-wing groups could become much smoother if each group approaches claims made by another with curiosity: “Could you please explain why you think Brexit is a good idea?”. Probe, question, and never become complacent about how much you understand something. Remember that if you ever you think you understand a topic/argument/mechanism sufficiently, you are probably wrong and can go deeper.

2. The Feynman Technique

This is a technique derived from Richard Feynman, the Nobel Prize winning physicist, when he was a student at Princeton.

Feynman believed that the ultimate test of how well you understood something, was your ability to explain/teach it to a child.

First, identify the topic you want to learn.

Write down everything you know about that topic in plain, simple language.

Second, write as if you were teaching it to a child.

The Feynman technique recommends that you write down everything you know about the topic, as though you are explaining it to a child.

Third, identify knowledge gaps.

Identify the gaps and the rough edges in your knowledge as you look back on your notes. Fill in those gaps.

Finally, tell a story concisely.

Spin your tale using analogies, anecdotes and other tools. Be concise. Be engaging.

3. The Protégé Effect: Learning by Teaching

The protégé effect is a psychological phenomenon where teaching, pretending to teach, or preparing to teach a subject to others, helps a person absorb that subject in greater detail.

When we teach, we learn.
— Seneca

Try teaching, or even prepare to teach a subject you want to learn thoroughly. Do you want to learn about distributed systems? Prepare teaching material on it! Create a presentation on distributed systems, with diagrams.