I haven't found a name for this fallacy, and perhaps it isn't one, but I would describe it as over-application of a general rule while disregarding specific information to the contrary.

but here are three examples:

"You want to lower the speed limit on this road to 40kph during business hours? So you think driving at night is safer?"

(Ignoring the specific information that this road is particularly dangerous during business hours)

"Of course he had a motive: most female homicide victims are killed by their ex-lovers"

(Ignoring evidence that the accused was not angry, was not violent, and was still on friendly terms with the victim)

"I'm not driving your car, 90% of that model die within the first 6 months"

(Ignoring the fact that this car is 2 years old, so it's obviously in the 10%).

I'm assuming in these cases that the speaker is aware of the specific information and is choosing to ignore it (or, more charitably, is unaware of its relevance).


Is it possibly just an example of a badly applied "argument by analogy"?

  • For instance, take: "Of course he had a motive: most female homicide victims are killed by their ex-lovers" This ignores the fact that many ex-lovers are not murderers. – Five σ Apr 26 '15 at 12:50
  • A weak analogy is generally where you assume that X is like Y, and then proceed to attach a property to Y. You then assume that since X and Y are similar, X also has property P. The problem is that X and Y aren't too similar in the first place. – Five σ Apr 26 '15 at 12:54
  • I think the idea that these claims ignore specific facts after introducing a general fact, is more compelling than the idea that they are "weak analogies". – Five σ Apr 26 '15 at 13:03

This is an accident (also known as "destroying the exception").

An accident is a deductive fallacy in which one ignores the existence of exceptions, and thus obtains a faulty conclusion.

As Ryder notes, the first one isn't a fallacy, as stated, but a slightly loaded question. The appropriate answer is "Yes, in the case of this road, driving at night is safer than driving during business hours".

For the other two, they are good examples of an accident. The example given on wikipedia is another one, which concludes that surgeons are criminals because cutting people with knives is a criminal act and surgeons cut people with knives... ignoring that there are exceptions in which cutting a person with a knife is not a criminal act.

| improve this answer | |

Understating the evidence

Philosopher Paul Draper has coined a fallacy known as "The fallacy of understated evidence".

Although its intended area of application is the philosophy of religion, I don't see why it can't be applied to other domains.

According to Draper, in the context of arguments for theism and against naturalism, proponents of a theistic argument are guilty of this fallacy if they “successfully identify some general fact F about a topic X that is antecedently more likely on theism than on naturalism, but ignore other more specific facts about X, facts that, given F, are more likely on naturalism than on theism.”

Here is an example of an argument that understates the evidence:

People have religious experiences of God, this is evidence for the existence of God.

Although the argument offers the general fact that people have religious experiences, it ignores the specific fact that many people never have religious experiences. Of those who do have these experiences, many already hold a prior belief in God.

The Ludic fallacy?

Nassim Taleb coined this term in his book The Black Swan. He defines it as basing studies of chance on the narrow world of games and dice.

The gynecologists example:

Person 1: Since about half the people in the world are female, the chances of the next person to walk out that door being female is about 50/50.

Person 2: Do you realize that is the door to Dr. Vulvastein, the gynecologist?

The gist of it is that reality involves a multitude of complex and concomitant variables, naive statistical models are not constructed to recognise this.

| improve this answer | |
  • Thank you, these are two very relevant fallacies that seem to apply in specific domains. I think the "accident" fallacy below nails it more generally? – Steve Bennett Apr 26 '15 at 13:40

FallacyFiles lists your last example almost explicitly as an unrepresentative sample. This would apply as well to the second example, but for the first I'm not sure this is really the same thing. I'd describe it (in the way you characterize) as suppressed evidence combined with an unrepresentative sample.

On the face of it, I'd simply answer "yes, driving at night is safer". To read it any other way obligates reading a rhetorical jibe into the question. Better to read it as an honest question, rather than assume the worst.

| improve this answer | |
  • I think it's almost the "unrepresentative sample" in reverse, isn't it? To use the marble example, I'm talking about arguments of the form "60% of the marbles in this bag are black, therefore 6 of the 10 marbles I pulled out are black" - even where is evidence suggesting otherwise. – Steve Bennett Apr 26 '15 at 12:40
  • For example #3 I'd qualify and say it's likely a better example of a combined suppression of evidence with the unrepresentative sample. For the other two Glen O is certainly right– I wasn't keenly aware of the Accident fallacy before, which is the much better fit (though I've known it as a "sweeping generalization"). Now you could characterize #3 as an Accident anyway, but that's just in the nature of the fallacy's definition: "there is no one fallacy of "accident", but a number of distinct fallacies have been discussed under that name." (source fallacyfiles.org/accident.html) – Ryder Apr 27 '15 at 9:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.