What does it mean that teaching for integration?

Explaining your knowledge to someone else forces you to integrate it.

Example: You have spent months studying cognitive biases, decision theory, and epistemology. You feel confident — you can recognize confirmation bias, you understand Bayesian updating, you know the difference between epistemic and instrumental rationality. Then a colleague asks you to explain how these ideas fit together. Not individually. Together. You open your mouth and discover that your knowledge exists in three separate compartments that have never been connected. You can explain each topic in isolation, but when pressed to show how biases relate to updating which relates to rationality, you stammer. The knowledge is present but unintegrated. The act of trying to teach it — to produce a coherent explanation for someone who does not already understand — reveals exactly where the connections are missing. And the effort to construct those connections in real time, under the pressure of an audience, forces integration that months of private study did not.

Try this: Choose a topic you have studied from at least two different angles — perhaps a concept you have encountered in multiple books, courses, or fields. Now explain it to someone as a single, coherent account. This can be a conversation, a written explanation, or even a voice memo addressed to a specific person. The rules: (1) You must connect the different sources into one unified explanation, not present them sequentially. (2) You must make it understandable to someone who knows nothing about the topic. (3) You must not use jargon without defining it. After the explanation, write a brief reflection: Where did you get stuck? Where did you discover gaps? Where did you find connections you had not noticed before? Those gaps and connections are the integration work that teaching forced you to do.