What It Means to Know Something

Imagine you glance at the clock on the kitchen wall. It reads 3:00, and it really is 3:00. Do you know what time it is? Most people would say yes — obviously. But now suppose the clock stopped working exactly twelve hours ago, and you happened to look at it at the one moment its frozen hands were correct. You believed it was 3:00. It was 3:00. You had a reason — you looked at a clock. And yet something feels off about saying you knew.

This little puzzle is what philosophers fight about when they ask what it means to know something. For a long time, going back to Plato, the standard answer was that knowledge has three parts. First, you have to believe the thing. You can't know that it's raining if you don't believe it's raining. Second, the thing has to actually be true. You can't know something false, no matter how confident you are. Third, your belief has to be backed up by good reasons — what philosophers call justification. A lucky guess doesn't count, even if the guess turns out right. Put these together and you get the classic definition: knowledge is justified true belief.

For about two thousand years, this seemed solid. Then in 1963, a philosopher named Edmund Gettier published a three-page paper that broke it. His move was simple. He described cases — now called Gettier cases — where someone has a belief that is true, and they have what looks like a perfectly good reason to hold it, but the reason and the truth line up only by accident. The stopped-clock example is one. You had a justified true belief, but your justification (the clock) was disconnected from what made the belief true (the actual time). It feels wrong to call that knowledge, because you were essentially lucky.

This matters more than it might seem. Once you notice the gap, you start seeing it everywhere. A weather app tells you it will rain tomorrow. It does rain. But suppose the app was glitching and randomly displayed "rain" for every city that day — you got the right answer for the wrong reason. A friend tells you a fact they learned from a website that turned out to be unreliable, even though, in this one case, the website happened to be correct. In each case, three boxes are checked: belief, truth, justification. And yet calling it knowledge feels like cheating.

Philosophers have spent the decades since Gettier trying to patch the definition. Some say we need a fourth condition: your reasons can't depend on any false step along the way. Others say justification has to be connected to truth in a reliable way, not just a lucky one. Still others argue the whole project is misguided and that "knowledge" is fuzzier than the three-part formula suggests.

What's worth taking from this is not a final answer but a sharper question. When you say you know something, you are claiming more than that you believe it, and more than that it happens to be true. You are claiming your belief is tied to the truth in the right way — not by accident, not by a broken clock. Figuring out what that right way looks like is harder than it sounds, and that difficulty is one of the things philosophy is actually for.