If you can't explain something in simple terms, you don't understand it (2017)

The claim that “if you can’t explain something in simple terms, you don’t really understand it” provokes sharp disagreement among technologists, educators, and scientists. Some see the ability to strip a concept to its essentials for a specific audience as evidence of deep understanding and rigorous thinking, while others argue that communication skill is a separate talent and many domains — from quantum mechanics to Kubernetes — depend on large amounts of prerequisite knowledge that cannot be compressed without distortion. The thread explores nuances such as reductionism’s limits, the role of the learner’s effort, how analogies can both illuminate and mislead, and what this all means for teaching, expert authority, and even how AI systems “explain” things.

Debating the aphorism and “understanding”

  • Many see “if you can’t explain it simply…” as a useful heuristic: deep understanding includes being able to decompose and reframe ideas.
  • Others stress a logical nuance: being able to explain simply is evidence of understanding, but not the only possible evidence.
  • Several argue that performance or rote use (e.g., calculating, driving, coding) doesn’t equal conceptual understanding.

Audience, simplicity, and pedagogy

  • “Simple” is repeatedly said to be audience-relative: a freshman, a layperson, a 5‑year‑old, and an expert peer need different explanations.
  • Good explanation often means starting from shared background and building stepwise; simple ≠ short or slogan-like.
  • Examples from math (compactness, NP, epsilon–delta, topology) show that definitions can be formally “simple” but pedagogically opaque.
  • Simple explanations may require long, layered, example-rich narratives (e.g., popular math/physics explainers).

Communication skill vs subject mastery

  • Many insist explaining clearly is a separate skill from understanding; plenty of strong practitioners are poor teachers.
  • Others counter that for true mastery, you eventually must also be able to externalize and structure your knowledge for others (or for yourself).

Irreducible complexity and limits of reduction

  • Some claim not all complexity is reducible without distortion (“computational irreducibility,” emergent phenomena, quantum mechanics).
  • Feynman’s difficulty explaining magnetism at a lay level is cited both as:
    • evidence we don’t “really” understand certain phenomena, and
    • a counterexample showing the aphorism overreaches.

Concrete examples and oversimplification risks

  • Multiple examples used to probe “simple explanation”: NP vs NP-hard, Heartbleed, Rowhammer, buffer overflows, Kubernetes/DevOps, calorie deficit and weight loss, Burrows–Wheeler transform.
  • Threads often converge on: you can give an intuitive core, but aggressive simplification can become wrong or misleading.
  • Distinction raised between analogy (can feel like understanding) and real, operational grasp.

Learner responsibility and misuse of the quote

  • Some complain the saying gets weaponized by people unwilling to do the work, blaming experts instead of their own lack of effort.
  • Others note the opposite misuse: experts blaming “lazy audiences” rather than examining their own weak explanations.
  • Several emphasize that understanding always requires work on the learner’s side; no explanation can fully substitute for that.

AI and LLMs

  • Discussion touches on LLMs “not really understanding,” but being good at simple explanations and Feynman-style breakdowns.
  • Concern that AI-driven simplification (e.g., endless summarization) may rob learners of desirable difficulty needed for real understanding.