One of the most important forms of practical reasoning is determining causes and effects. Being able to understand and explain the past, and predict and achieve our ends in the future, requires causal reasoning.
When we reason poorly in identifying such causal relationships we commit the questionable cause fallacy.
There are uncountable possible ways one might arrive at an unjustified causal belief, but there are some that are common enough to be worth mentioning. (These do overlap; I’m not claiming them to be completely different types of errors.)
* One is to assume that correlation is causation. For instance, someone might point out that the divorce rate has been higher in recent decades, note that television has been a big part of people’s lives ever since it was invented several decades ago, and claim that there’s therefore something about television that has weakened marriages and led to more divorces.
Well maybe. But there are a huge number of other things that have existed for the last several decades-the wide availability of birth control, racial integration, rock and roll music (and other modern music like hip hop), the Frisbee, etc.-and if all it took was correlation, then one would be just as warranted in claiming that one or all of these causes a higher divorce rate. So clearly it takes more to justify a causal claim than simply that certain things have been correlated with each other.
* A second is to assume that because one event followed another event, the first event caused the second.
There’s even a fancy Latin name for this one: “post hoc ergo propter hoc” (“after this therefore because of this”), or simply the “post hoc” fallacy.
Medical patients, and “alternative” medicine practitioners commit this one left and right. “I took this herb [or prayed, or had a nurse wave her hands over me without touching me to ‘balance my energy fields,’ or had my ear candled, or went to the Philippines and had ‘psychic surgery,’ etc.] and then I got better. So obviously this herb is effective in curing what I had.”
Indeed, people’s beliefs in things like this can be unshakeable. “I don’t have to doubt it or question it. I don’t need more evidence. It actually happened to me. What possible better evidence could there by? I experienced it!”
No, see, you experienced A and then B. You didn’t “experience” A causing B. That’s something you added on as an interpretation.
Medical and scientific researchers learned long ago about something called the “placebo effect,” wherein if people expect something to make them better, then that expectation alone improves their condition above chance levels. For example, if people are told they’re being given an amazing new drug that lowers blood pressure, then the blood pressure of the average person taking it will decrease, even if it’s all a lie and they’re being given plain water or a sugar pill.
So if you expected the herb [or the prayer or the “therapeutic touch” or whatever] to help, it may have been that expectation rather than the herb that caused your improvement. And if properly conducted medical experiments show no connection between that herb and the condition you had, that constitutes vastly better evidence than your experience.
Though it doesn’t have to be just the placebo effect. Remember, in addition to the herb, you were probably getting other treatment as well. You don’t know which of them made you better.
And for that matter, time itself is a kind of treatment. Sometimes we get better for no discernible reason other than that the body had enough time to heal.
I have a doctor friend who chuckles whenever he sees claims that an impressive 30% or whatever of patients suffering from low back pain reported feeling noticeably better within such-and-such amount of time after taking this vitamin or that herb or having chiropractic treatment, etc. “You know what has the identical success rate over that period of time?” he’ll ask me, “Doing nothing whatsoever. Just resting and taking it easy.”
* A third is to ignore the possibility that rather than A causing B, maybe something else caused them both.
A trivial example would be if, say, a child was scared to go to a hospital because the people in his life who’d died all seemed to have gone to the hospital before doing so.
Well, he’s right there’s a correlation. People who go to the hospital do die more quickly on average than people who don’t. But it’s not going to a hospital that causes death. (Not necessarily anyway. Certainly there have been times and places throughout history where you were at greater risk if you got the currently practiced medical treatment than if you didn’t. Think of bloodletting for instance.) Being unhealthy causes people to decide to go the hospital, and being unhealthy causes people to die.
If unhealthy people refrained from going to the hospital, all that would happen is they’d be at even greater risk of dying.
If you’re a sports fan, there’s a wonderful example of just this fallacy that’s been around for decades, and really only recently have most coaches and informed fans figured out it’s bogus. (“Most” being key; in some circles, this is still treated as common knowledge.)
For the longest time, in football it was said that a strong running game was more important than a strong passing game. Or really just that being willing to call a lot of running plays gives you the best chance to win. On what evidence? People noted that the teams that had the highest ratio of run plays to pass plays won the majority of their games, while teams that had the most pass plays tended to do poorly.
Indeed, if you look at the statistics after this weekend’s games, you’ll likely note a surprising correlation between quarterbacks passing for a lot of yards and losing. Check out the ten quarterbacks who threw for the most yards, and likely fewer than half won their games. So not only, it seems, is it better to run than to pass, it’s better to run than to pass successfully.
Have you spotted the flaw yet? Again it’s a case of A and B having a common cause rather than one of them causing the other.
In football, running plays on average use up more time, so teams that are ahead tend to run in order to give their opponents less time to catch up, while teams that are behind tend to pass so as to do what they’re going to do-succeed or fail-quickly to save time for additional opportunities. So the team in the lead wants to run out the clock; the trailing team has the opposite goal.
Given that, we see that being in the lead “causes” more running plays. But of course it also “causes” more victories. Obviously teams that are ahead 21-14 going into the 4th quarter win more often than teams that are behind 21-14.
So it’s not that running makes you more likely to win. It’s that being ahead makes you both more likely to run and more likely to win.