The Deceptive Familiarity of Understanding
We use the word constantly. "I understand," we say, in conversation, in classrooms, in arguments. But pressed to explain what understanding actually is — what makes it different from merely knowing, or from being able to repeat something back — the concept becomes elusive. This isn't a trivial puzzle. It sits at the heart of education, cognitive science, philosophy of mind, and even debates about artificial intelligence.
Knowledge vs. Understanding: Is There a Difference?
Consider two students who both correctly answer every question on a physics exam. One has memorized the formulas and solved enough similar problems to recognize patterns. The other can derive the formulas from first principles, explain why they work, and apply them to novel situations they've never seen before. Both know the material — but we'd likely say only the second truly understands it.
This distinction matters. Knowledge can be stored and retrieved. Understanding is generative — it lets you do things with knowledge that you haven't been explicitly taught. You can explain, adapt, predict, troubleshoot, and create on the basis of understanding. With mere knowledge, you can only reproduce.
Philosophical Takes on Understanding
Philosophers have approached this from several angles:
- The coherentist view: You understand something when it fits coherently into a web of other things you believe and know. Understanding is about integration, not isolated facts.
- The ability view (Gilbert Ryle): Understanding is a cluster of abilities — being able to use information appropriately, to recognize when it applies and when it doesn't. It's "knowing how" as much as "knowing that."
- The phenomenological view: Understanding involves a subjective experience — the "aha" moment, the felt sense of things clicking into place. This is what learners and teachers often mean when they talk about genuine understanding as opposed to surface mastery.
The Feynman Technique as a Test
Physicist Richard Feynman had a practical test for understanding: if you can't explain something clearly to someone with no background in the subject, you don't really understand it. You might know the terminology, but terminology can be a mask for confusion. True understanding expresses itself in simple, clear language because it doesn't need jargon to hold together.
This insight gave rise to the "Feynman Technique" as a learning method: take a concept, try to explain it in plain language, identify where your explanation breaks down, and go back to the source material to fix the gaps. The breakdowns are where the real learning happens.
Does AI "Understand" Anything?
This question has sharpened considerably in recent years. AI systems can now produce text that explains concepts accurately, passes exams, and answers questions in ways that look exactly like understanding. But many researchers and philosophers argue that this is sophisticated pattern-matching without genuine comprehension.
John Searle's famous Chinese Room thought experiment illustrates the issue: imagine a person in a room who receives Chinese characters, follows rules to manipulate and return them, and produces correct Chinese responses — without understanding a single word of Chinese. The system behaves as if it understands, but does it? This debate remains genuinely unresolved and touches on deep questions about consciousness, meaning, and mind.
Why This Matters for Learning
Taking understanding seriously — as something distinct from and more valuable than memorization — has real implications for how we learn and teach. It suggests that:
- Testing recall is not the same as testing understanding.
- The best learning involves making connections across domains, not mastering isolated units.
- Confusion, properly engaged, is a productive part of the learning process — not a sign of failure.
- Being able to teach something is one of the highest tests of your own understanding.
Understanding understanding, it turns out, is one of the more useful things you can do with your mind.