Fallacy Fallacy

The Fallacy Fallacy:

Also called

Claiming that a position must be false because the argument used to get to that position is invalid or used a fallacy. It may sound like a rational thing to do since by definition a fallacious argument makes no sense, and this rule may seem like a mindscrewy special case, but...

Tom: All cats are animals. Ginger is an animal. This means Ginger is a cat.
Bill: Ah, you just committed the "affirming the consequent" logical fallacy. Sorry, you are wrong, which means that Ginger is not a cat.

Bill's rebuttal is an argument from fallacy, because Ginger may very well be a cat; we just can't assume so from Tom's argument.

In other words, pointing out somebody's fallacy is not fallacious in itself (you're doing it right), but using this as "proof" that their claim is false is the Fallacy Fallacy. Somebody arguing their point badly doesn't automatically mean they are wrong. The best you can say is that they have not convincingly supported it. This also applies to the Fallacy Fallacy itself: Bill's argument is a fallacy, but it would be the same fallacy to conclude that Ginger is a cat because of that, since Tom's only "proof" is not a valid argument.

Another excellent example of how a false argument combined with a true conclusion: in medicine, pressure around the brain can cause severe headaches. Ancient surgeons assumed that it must be demons in the patient's head causing the pain, and that exposing them to light would kill them or drive them out; therefore, they drilled holes in the patient's skull. The end result relieved the pressure and actually did cure the headaches, even though their reasoning was entirely faulty.

An argument using fallacious reasoning is capable of being consequentially correct. In logic, "invalid" (fallacious argument) and "false" are not synonymous (See Sound/Valid/True for a more complete explanation of this. There are reasons why extensive Critical Thinking courses exist.) This is related to how logical argument is used as a tool rather than as a fact-in-itself, and that logical validity can sometimes be surpassed by an objective scientific fact.

See also Right for the Wrong Reasons, Dumbass Has a Point, and Don't Shoot the Message.


  • Any Straw Vulcan character is bound to be written as if "illogical" is a synonym for "wrong."
  • A good many theories about the world over the years including many scientific ones are mostly Wild Mass Guessing for their time that lead to the right conclusions but for the wrong reasons. Freud's theories, for example, are sometimes useful but are based on dated inaccurate knowledge of the mind.
  • One of the standard examples of this is Continental Drift: Alfred Wegener was right in that the continents moved, but the process he postulated was dismissed by geologists as being absurd (and they were right). It was the discovery of the mid-ocean ridges and the trenches in the 1950s and 1960s that finally demonstrated the how of plate tectonics.
  • Of course, a proposition may fallaciously be declared correct due to an argument against it fallaciously declaring it incorrect due to a particular argument for it being fallacious, thus committing the Fallacy Fallacy Fallacy.
  • One specific Fallacy Fallacy is based off of the Appeal to Ridicule. This is very common in political debates, wherein if an individual ridicules some position without backing up the ridicule, the opposing side will assume that the Appeal to Ridicule was made because the person has no actual argument to make. Instead, the person who made the argument was just trying to be funny, or was just taking some time to enjoy disparaging the opposition. Or the argument was so transparently ridiculous they didn't think it was worth their time to discredit it (who wants to be the politician caught on camera spending half an hour to explain that they can't disprove the existence of an orbiting teapot?). And of course, even if the person was engaging in a fallacy, it doesn't say anything about others who share their point, and may very well be able to back up their claims.
  • Kinda inverting: Martin Gardner, the math popularizer, coined the Ycallaf, which is an argument that *looks* like a fallacy but is nevertheless true. (Beloved example: The otherwise fallacious 0=1-1+1-1+...=1-(-1+1-1+)...=1 argument is actually valid when properly interpreted in the context of knot theory.) note 
  • A beautiful example of the Fallacy Fallacy comes up every time an agency, government, or corporation involved in a controversial issue conducts an internal review or investigation and then publishes results which agree with their desired position. The doubters will immediately resort to an impugning motives form of ad hominem; "Of course that's what they would say! They've got a vested interest!" This may be true, but one cannot logically conclude that the impure motives mean the internal review was falsified. However, in turn, this does not mean there was no cover-up at all. Logic cannot tell us what is true in this case, so we have to switch to other modes of thought to come to a probabilistic conclusion about whether "they" are lying.
    Alice: The environmental survey says there will be no damage to caribou populations if we open this area to mining.
    Bob: Of course that's what those greedy bastards at Strippit Mining Enterprises would say. They're just publishing research which supports their point of view.
    Alice: Hold on; you're just dismissing the report because the company did the research without even examining the report.
    Bob: And you're uncritically accepting their authority when they have a clear interest in the conclusion.
    Charlie: So basically...you're both talking out of your rear ends.
    • Alternate hypothesis analysis shows that an organization coming to a conclusion that suits their vested interest *is* evidence that the conclusion is true—just very weak evidence. Consider: if the mining would harm caribou, what would we expect Strippit Mining to say? That it wouldn't harm caribou,of course. But, suppose the mining would in fact not harm the caribou—what would we expect Strippit to say? Why, that the mining would not harm the caribou. Since one would expect Strippit to say the same thing whether or not it is true, their statement doesn't seem to be evidence one way or the other—and definitely not evidence that their finding is untrue. However, if the mining would harm caribou, there is at least some chance Strippit would admit the harm. Organizations, like people, are sometimes honest even when it is against their best interests. But if the mining would not harm the caribou, it is hard to imagine Strippit lying and saying it would, a lie which would be against their own interests. Therefore, their claim of harmlessness is weak evidence of harmlessness, based on how likely Strippit is to be honest against their own interest. (And Strippit reporting that their mining plan would harm caribou is rather strong evidence that it would. Organizations don't usually lie to harm themselves.)
      • Additionally, the value we place on Strippit's evidence in part depends on context. If, for example, mining is well-regulated and companies are investigated by thorough, incorruptible officials who are likely to find any malfeasance and fines for deceit are punishingly high, we can consider Strippit's real vested interest is in being honest. The same might be true without laws in an environment where people internalize social responsibility, such as if Strippit's corporate culture consisted of people from the land to be mined and they valued environmentally responsible mining.
      • An example of this sort of policing is in clinical trials, where a pharmaceutical company really is interested in catching common adverse effects during the Phase III trials (before releasing the drug to market), as there are whole bevies of lawyers waiting to discover a drug has adverse effects that were swept under the rug. More problematic are rare side effects discovered during Phase IV (post-release surveillance), as the population taking the drug explodes and new events may be found that didn't show up in the first few hundred trial participants.
    • Note also that this sort of bias need not involve deliberate lying—people will interpret ambiguous evidence as favoring the conclusion they desire or already believe. That is, assuming the evidence is ambiguous, we would expect the Ban All Mining And Especially Strippit activist group to report that the mining would harm the caribou - and not necessarily through any lying or even lack of diligence on their part.
    • And of course if either Strippit or Ban All Mining offers actual proof, this is not disproven by their own self-interest or preexisting ideology.