Oct. 19, 2026
Question: How do you know if/when you understand something?
Is a flash of insight enough?
Sometimes it intuitively feels like we understand a concept, perhaps especially when we experience a sudden burst of clarity.
But that’s not understanding…that’s just feeling like you possess understanding.
And we know that feeling can be wrong, because we periodically discover we’ve been mistaken about a belief that previously felt intuitively clear.
Is it that we can correctly answer questions about the topic?
It does seem true that if you understand something, you should be able to answer questions about it. This also aligns well with our intuitions based on educational systems where knowledge is evaluated in this way.
But how do we know that we’re being asked the right questions in the first place? Additionally, how do we know that answers deemed “correct” are reflective of true understanding?
Consider this question, which is exactly the sort of thing that is often posed to college students in Intro Stats courses:
To compare the means of two independent groups, which statistical test would be appropriate?
A. t-test
B. ANOVA
C. Regression
D. Correlation
A student who selects “t-test” as their response might get a good grade on the assignment. And if they answer many such questions, they might get a good grade in the class.
If they then complete multiple statistics courses, they might come to believe they understand statistics well.
The problem, however, is that all four answer choices would have been correct! These are different ways of representing the same concept (see this video explanation).
In this case, the instructor’s own lack of understanding trickles down to the students. For them, “learning” just amounts to a sophisticated form of pattern-matching, where the link between a prompt (“comparing means of two independent groups”) and a response (“t-test”) is memorized.
As shown above, this form of pattern-matching can happen without there being any deeper understanding of what is actually going on.
So, clearly, being able to get the “right answer” to questions on the topic does not necessarily signal understanding.
Indeed, having completed trainings, certifications, courses, or even entire degree programs does not necessarily indicate that the learner possesses understanding of the relevant subject.
Is it that others describe us as possessing understanding?
For example, if everyone around you gushes about how well you grasp a particular subject, does that make it true? What if recognized authorities on the subject formally induct you into their ranks based on long and careful judgment?
This could certainly make us feel like we understand, at least if we’re inclined to take others’ word for it.
But does it really make sense to equate understanding (a psychological state inside your head) with social perception (a psychological state inside others’ heads)?
That seems wrong by definition.
Maybe if we don’t believe that “objective understanding” is even possible, because of the natural limits of how our brains work, then we would be satisfied to conceptualize understanding as a purely social construct?
In fairness, although they’re limited, I do think each of these could be useful indicators of what we call understanding, especially when taken together:
- You perceive that you grasp a concept.
- Others agree, including others who are also thought to possess understanding of the concept.
- You successfully pass tests intended to measure understanding.
But even with all these criteria taken together, it’s not difficult to imagine a situation where you still don’t really understand, because no one involved really understands.
(This is not that far-fetched. It’s arguably the case for any system so complex that we can’t make high-accurate predictions.)
It everyone misunderstands how things work, and that misunderstanding is also reflected in the tests/questions, then achieving all these criteria still doesn’t grant something like “objective” or “true understanding.”
So maybe we never truly know whether we understand…
In which case, “How do you know if you understand something?” is not a meaningful question to pose.
Perhaps part of the issue is that understanding is not actually a binary state in your mind.
Maybe the question is not “Do you understand?” but rather “How well do you understand?”
