The fallacy fallacy fallacy is not a common one because it requires two unlikely interlocutors. One side needs to know that the fallacy fallacy exists. The other side needs to call out fallacies, and call them out correctly (i.e. not committing the fallacy fallacy).
But I personally see fallacy fallacy fallacies all the time, because I do call out fallacies when I see them and call them out correctly (in my estimation). I believe most rational people should try to do this; the idea of fallacies exists for a reason.
On the other hand, a lot of people do not even understand what a fallacy actually is, and therein lies the problem: people who do not know what a fallacy is, use the fallacy fallacy as a tactic to get the other interlocutor to shut up about fallacies.
Let’s start from the bottom.
Fallacy
A fallacy is an error in reasoning, plain and simple. In particular, certain errors in reasoning follow patterns that are so common that they get names, for example ad hominem.
An argument that contains a fallacy is invalid, that is, even if the premises of the argument were true, the conclusion is not necessarily true. However, that doesn’t mean the conclusion is necessarily false.
I find it easier to flip the problem around: if an argument is valid and the premises are true, then the conclusion has to be necessarily true. Or another way to put it: if the argument is sound.
Fallacy fallacy
If the argument is sound, then the conclusion is necessarily true. But if the argument is unsound, does that mean the conclusion is necessarily false? No, it’s not necessarily true.
If one concludes that the conclusion is necessarily false, that would be an inverse error fallacy, and it’s called a fallacy fallacy.
If an argument is unsound we cannot be certain of the conclusion: it’s possibly false, or unknown, but it’s possibly true as well. All we know is that it’s not necessarily true.
Fallacy fallacy fallacy
If an argument X
contains a fallacy, it’s invalid, and therefore the conclusion is not necessarily true.
If interlocutor B
claims that because argument X
contains a fallacy, the conclusion must be necessarily false, that would be a fallacy fallacy.
But if interlocutor B
claims that because argument X
contains a fallacy, the conclusion is not necessarily true, that’s not a fallacy fallacy. Thus this argument Y
is valid.
Then if interlocutor A
claims argument Y
is a fallacy fallacy, he would be wrong. This argument Z is a fallacy fallacy fallacy. If interlocutor B
claimed the conclusion must be necessarily false, that would be a fallacy fallacy, but he claimed it’s not necessarily true, which is different.
Argument X has a fallacy. Argument Y is valid for pointing to the correct conclusion of X. Argument Z is invalid for claiming Y has a fallacy.
Claiming that it’s a fallacy to conclude that the conclusion of a fallacious argument is not necessarily true is a fallacy fallacy fallacy.
Is that clear? Consider this argument: Eugene says Socrates is mortal, Eugene is an authority, therefore Socrates is mortal. That is clearly a fallacious argument, but it would be a mistake to assume the conclusion is false: Socrates is mortal after all. Although it’s not a fallacy to point out the conclusion of this particular argument is not necessarily true, given that we arrived at it through a fallacious argument.
Claiming that it’s not necessarily true that Socrates is mortal is not a fallacy fallacy, and claiming that it is would be a fallacy fallacy fallacy.
False or untrue?
I believe the reason why so many people have issues distinguishing correct calls to a fallacy fallacy versus incorrect calls, is that most people don’t see the difference between necessarily false, and not necessarily true.
Isn’t false
the opposite of true
? Not necessarily.
The problem is that most people just assume bivalent logic, but there’s many kinds of logic. In classic bivalent logic ¬true
is false
, sure, but not in modal logic.
In modal logic the negation of necessarily true (¬□⊤
), is possibly false (◇⊥
), not false.
It turns out something as simple as the material conditional (p ⇒ q
) gets really complex really fast if you don’t assume bivalent logic. The Stanford Encyclopedia of Philosophy has an entire entry devoted to the differences in different logics (modal, three-valued, etc.): The Logic of Conditionals.
What we are looking for is the negation of a material conditional, which in bivalent logic makes no sense. Consider a simple example: if a man is rich (p
), then he is smart (q
), in bivalent logic the negation is p ∧ ¬q
: men are rich and dumb. Clearly classical bivalent logic is not the be-all and end-all.
What we are trying to say is p ⇏ q
, that is: even if the premises of the argument are true, the conclusion is not necessarily true. The fallacy fallacy fallacy is confusing that with p ⇒ ¬q
(the conclusion is false).
For a deeper exploration of the difference, I wrote: First principles of logic and not-guilty is not the same as innocent.
Conclusion
Don’t let people discourage you from pointing out fallacies.
There’s a good reason why the notion of fallacies exists in the first place: they invalidate arguments (which is not to say they make the conclusion false).
Of course people don’t like to be caught committing fallacies, if they have the fallacy fallacy card, many will wrongly play it. If you never claimed the conclusion was necessarily false, then return the card back to them with the fallacy fallacy fallacy.