Incorrect Cause
From SkepticWiki
Contents |
[edit] Definition
Incorrect Cause is an informal logical fallacy (also known by the Latin non causa pro causa, "not a cause [taken] for a cause") where a participant invalidly assumes a causal relationship between two observations where it is not supported. There are several different types of incorrect causes, depending upon the specific observations.
[edit] Example
Example 1:
- Antagonist: I wore my blue sundress to the golf course today, and I had the best round of my life! Obviously, my blue sundress is lucky.
Example 2:
- Antagonist: I ate lots of oranges yesterday, and today my cold is gone. Oranges cure colds.
Example 3:
- Antagonist: In the United States, gun ownership is among the highest in the world -- and so are the murder rates. Obviously, owning guns must increase the crime rate.
Example 4:
- Antagonist: Have you ever noticed that people with lung cancer tend to be heavy smokers? Obviously, cancer causes smoking.
Example 5:
- Antagonist: Fifty percent of cardiac arrests in hospitals occur in intensive care units. Therefore, ICUs cause heart attacks.
Example 6:
- Antagonist (Homer): Not a bear in sight. The "Bear Patrol" must be working like a charm!
- Protagonist (Lisa): That's specious reasoning, Dad.
- Antagonist (Homer): Thank you, dear.
- Protagonist (Lisa): By your logic I could claim that this rock keeps tigers away.
- Antagonist (Homer): Oh, how does it work?
- Protagonist (Lisa): It doesn't work.
- Antagonist (Homer): Uh-huh.
- Protagonist (Lisa): It's just a stupid rock. But I don't see any tigers around, do you?
- Antagonist (Homer): (pause) Lisa, I want to buy your rock.
[From the Simpsons (season 7, "Much Apu about Nothing")]
[edit] Discussion
Philosophers and logicians have identified a number of subcategories of the fallacy of Incorrect Cause, many of which have their own name.
- complex cause: Selecting only one of a number of causes for a given effect and neglecting the others. For example, "the Challenger disaster was caused by cold weather, an act of God. It couldn't have been helped." While partially true, the design and construction of the O-rings used were inappropriate for cold weather launches. Had the O-rings been properly constructed, or had the flight been cancelled, the disaster could have been prevented.
- cum hoc ergo propter hoc: Literally, "with this, therefore because of this." Concluding that because two events happened together, one must cause the other, as in examples 1, 5 and 6 above.
- joint cause: Concluding that one thing caused another when they are really both caused by the same thing. For example, there is a strong correlation between the height of elementary school children and their ability to read, but this is not because reading better makes you get taller. Instead, as children age, their reading level and height both increase. Example 3 may be of this type, where both the murder rate and gun ownership rates may be caused by other aspects of the culture of the United States.
- post hoc ergo propter hoc: Literally, "after this, therefore because of this." Concluding that because one event preceeded another, it caused the other, as in example 2 above. (Although eating the oranges came before her cold getting better, it might simply have cleared up all by itself.)
- regression fallacy: "Regression to the mean" is a well-known statistical phenomenon where a measured extreme value is later measured to be much closer to average. For example, in order to be a championship sports team, a team will have to be both lucky and skilled; many times, a team that wins a championship one season will have a merely average season the next, because although they are equally skilled, they simply did not get the same lucky breaks. Assuming that something other than regression to the mean caused the decreased performance is the regression fallacy.
- Texas Sharpshooter fallacy: Assuming that any observable pattern in the data must have a cause. The name derives from the joke about a sharpshooter who achieved a reputation for incredible accuracy by firing first, and then drawing a target around the (randomly-located) bullet hole.
- wrong direction: Inferring a genuine causal link, but in the wrong direction, as in example 4 above.
[edit] Exceptions to the Rule
Any inference of cause is potentially fallacious, but most of science nevertheless hinges on the ability to infer causes "rationally." The key to analyzing an argument of this form is to consider the alternate hypotheses that could also produce the observed behavior. For example, if cancer doesn't cause smoking, what's the actual relationship? Is it likely that smoking causes cancer? Is it plausible that the observed relationship is simply a coincidence, with no real cause? Could they both be caused by some other, previously unsuspected joint cause? For example, could people with the genetic tendency to cancer also have a genetic tendency to smoking behavior?
