If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

Fallacies: Post Hoc Ergo Propter Hoc

In this video, Paul explains the post-hoc-ergo-propter-hoc fallacy. This is an informal fallacy committed when a person reasons that because one event happened after another event, the first event caused the second. He also discusses why it is sometimes hasty to conclude that your cat scratch caused your fever.

Speaker: Paul Henne, Duke University.

Want to join the conversation?

Video transcript

(intro music) Hello! I'm Paul Henne, and I'm a philosophy graduate student at Duke University. In this video, I'm gonna talk to you about an informal fallacy called "post hoc ergo propter hoc." In doing this, I'm also gonna tell you why it's sometimes wrong to conclude that a cat scratch caused your fever. But I'll get to this idea in a bit. Let's start with a problem. Suppose that one day, I decide to work from home, and my roommate Alex asks to use my bike, and of course I let him. So he takes my bike, and goes to work. Later that day, he returns home with my bike, only the chain is busted. I say, "Alex, you broke my bike! "You caused the chain to break! "You're responsible for this!" He rode my bike, and then it broke. It seems as though he caused it to break. So he should be responsible, right? But is this the correct conclusion for me to draw? Did Alex cause my bike to break? What do you think? I'd like to suggest that that's the wrong conclusion to draw. Well, it's wrong for me to conclude this without gathering more evidence that Alex was negligent while riding my bike. For it could be the case that the chain on my bike was defective in some way, and that whoever rode my bike that day, maybe me, would've returned home with a broken bike. In this scenario, I've committed the post hoc ergo propter hoc fallacy. "Post hoc ergo propter hoc" is a Latin phrase roughly meaning "after this; therefore, because of this." Here is the idea. It's a fallacy to argue that because one event (like my bike breaking) occurred after another event (like Alex riding it), the first event caused the second event. Let's represent this formally. Premise one: X happened after Y. Conclusion: Therefore, Y caused X. Returning to our example, my argument now looks like this. Premise one: My bike broke after Alex rode it. Conclusion: Therefore, Alex caused it to break. So, is this a valid argument? What do you think? As it is now, the argument is invalid. Here's a counterexample: the sun sometimes comes out after it rains, but that doesn't mean that the rain caused the sun to come out. But let's be more charitable to my invalid argument. Maybe we can add a few premises, implicit ones, to make it valid, and to make it seem more reasonable. Let's try this. Premise one: My bike broke after Alex rode it. Premise two: Usually, the person who rides the bike before it broke caused it to break. Conclusion: Therefore, Alex caused it to break. But this argument is now inductive, which is fine. However, it shows us that we need more evidence that Alex broke my bike in order to conclude that he broke it, as opposed to the bike just breaking, no matter what. So, let's consider two cases with more or less evidence. Case one: Suppose my other roommate, Aleah, saw Alex hitting my bike chain with a hammer, and then she told me about it. If this were the case, I would have better evidence, a better case for concluding that Alex broke my bike. So let's consider now case two. Suppose Aleah did not see Alex hitting my bike with a hammer. I can conclude, although hastily, that it's only likely Alex who broke my bike, though I'm still committing the fallacy in this case, and I need more evidence, like the testimony of my roommate. Let's take another stab at my argument. Premise one: My bike broke after Alex rode it. Premise two: If a bike breaks after someone rides it, the rider caused it to break. Conclusion: Therefore, Alex caused my bike to break. Is this argument valid? Yes! But wait: is it sound? Let's check. Is premise one true? Yes. Is premise two true? No! Plenty of bikes break without the rider causing it. So suppose Alex didn't act negligently at all. Like I said earlier, the bike could've been worn out. So no one acted negligently. It's just natural wear and tear of the bike. And the natural wear and tear caused the bike to break, not Alex. Here, premise two is false. Regardless of the exact argument representation I use, I need more evidence about this case for me to make my conclusion. If I don't have evidence that Alex acted negligently, I'll not have a good argument. I'll make a hasty conclusion, and I'll commit the post hoc fallacy. We can see now that this "after this; therefore, because of this"-reasoning is defective. A cat might've scratched you right before you got a fever, but that doesn't mean that the cat scratch caused your fever. It's perfectly possible that the flu that your roommate recently had was the actual cause for your fever. The sequence of the two events alone does not establish a causal relation between them. You probably have a lot of questions about this fallacy now, and there's a lot more to talk about, like causation, controlled experiments, possibility, and superstitions. But unfortunately, they'll have to wait for other videos. Subtitles by the Amara.org community