You argue that it would be wrong to stab my neighbor and take all their
stuff. I reply that you have an ugly face. I commit the “ad hominem” fallacy
because I’m attacking you, not your argument. So one thing you could do is
yell “OI, AD HOMINEM, NOT COOL.”
[...] What you need to do is go one step more and say “the ugliness of my
face has no bearing on moral judgments about whether it is okay to stab your
neighbor.”
But notice you could’ve just said that without yelling “ad hominem” first! In
fact, that’s how all fallacies work. If someone has actually committed a
fallacy, you can just point out their mistake directly without being a pedant
and finding a pat little name for all of their logical reasoning problems.
Fallacy names are useful for the same reason any term or technical vocab are useful.
‘But notice how you could’ve just you meant the quantity 1+1+1+1 without yelling “four” first! In fact, that’s how all ‘numbers’ work. If someone is actually using a quantity, you can just give that quantity directly without being a mathematician and finding a pat little name for all of their quantities used.′
Fallacy names are great for chunking something already understood. The problem is that most people who appeal to them don’t understand them, and therefore mis-use them. If they spoke in descriptive phrases rather than in jargon, there would be less of an illusion of transparency and people would be more likely to notice that there are discrepancies in usage.
For instance, most people don’t understand that not all personal attacks are ad hominem fallacies. The quotation encourages that particular mistake, inadvertently. So it indirectly provides evidence for its own thesis.
If you’re assuming that they won’t be punished if they convinced the other person, then that’s true. That would be a conflict of interest and hint at them starting with the bottom line.
If you don’t assume that, then it sounds like ad hominem combined with circular logic. Them being a murderer doesn’t mean their argument is wrong. In fact, since they’re living the conclusion, it’s evidence that they actually believe it, and thus that it’s write. Furthermore, them being a murderer is only bad if you already accept the conclusion that it’s not OK to kill the other person and take their stuff.
You can’t say that whenever they are a murderer or not has no relation to the argument they’re making, while you can say that for the face being ugly, though.
I voted your comment up because I agree that the vocabulary is useful for both the person committing the fallacy and (I think this is overlooked) for the person recognizing the fallacy.
However, I think the point of the original quote is probably that when someone points out a fallacy they are probably felling angry and want to insult their interlocutor.
It’s like when those stupid car buffs say “Hmmm...yeah, transmission fluid” when telling each other what they think is wrong rather than “It sounds like the part that changes the speed and torque with which the wheels turn with respect to the engine isn’t properly lubricated and able to have the right hydraulic pressure, so you should add some green oil product.”
That’s not even an example of the ad hominem fallacy.
“You have an ugly face, so you’re wrong” is ad hominem. “You have an ugly face” is not. It’s just a statement. Did the speaker imply the second part? Maybe… but probably not. It was probably just an insulting rejoinder.
Insults, i.e. “Attacking you, not your argument”, is not what ad hominem is. It’s a fallacy, remember? It’s no error in reasoning to call a person ugly. Only when you conclude from this that they are wrong do you commit the fallacy.
So:
A: It’s wrong to stab your neighbor and take their stuff. B: Your face is ugly. A: The ugliness of my face has no bearing on moral... B, interrupting: Didn’t say it does! Your face is still ugly!
Did the speaker imply the second part? Maybe… but probably not.
They did not logically entail it but they did conversationally implicate it (see CGEL, p. 33 and following, for the difference). As per Grice’s maxim of relation, people don’t normally bring up irrelevant information.
B, interrupting: Didn’t say it does!
At which point A would be justified in asking, “Why did you bring it up then?” And even if B had (tried to) explicitly cancel the pragmatic implicature (“It’s wrong to stab your neighbor and take their stuff”—”I won’t comment on that; on a totally unrelated note, your face is ugly”), A would still be justified in asking “Why did you change the topic?”
B here is violating Grice’s maxims. That’s the point. He’s not following the cooperative principle. He’s trying to insult A (perhaps because he is frustrated with the conversation). So applying Gricean reasoning to deduce B’s intended meaning is incorrect.
If A asks “why are you changing the subject?”, B’s answer would likely be something along the lines of “And your mother’s face is ugly too!”.
My point is that Grice’s maxims, useful though they are, do not fully capture how human conversation goes — most notably, those cases in which at least one party has a hostile or uncooperative attitude toward the other. People in such cases do get that they’re being insulted or whatever; A, as you portray him, comes off as simply bad at understanding non-literal meaning (or he is being intentionally obstructive/pedantic).
“You have an ugly face, so you’re wrong” is ad hominem. “You have an ugly face” is not. It’s just a statement. Did the speaker imply the second part? Maybe… but probably not.
I contest the empirical claim you are making about human behaviour. That reply in that context very nearly always constitutes arguing against the point the other is making. In particular, the example to which you are replying most definitely is an example of a fallacious ad hominem.
A: The ugliness of my face has no bearing on moral…
In common practice it does. The rules do change based on attractiveness. (Tangential.)
-- TychoCelchuuu on Reddit
Fallacy names are useful for the same reason any term or technical vocab are useful.
‘But notice how you could’ve just you meant the quantity 1+1+1+1 without yelling “four” first! In fact, that’s how all ‘numbers’ work. If someone is actually using a quantity, you can just give that quantity directly without being a mathematician and finding a pat little name for all of their quantities used.′
Fallacy names are great for chunking something already understood. The problem is that most people who appeal to them don’t understand them, and therefore mis-use them. If they spoke in descriptive phrases rather than in jargon, there would be less of an illusion of transparency and people would be more likely to notice that there are discrepancies in usage.
For instance, most people don’t understand that not all personal attacks are ad hominem fallacies. The quotation encourages that particular mistake, inadvertently. So it indirectly provides evidence for its own thesis.
Yeah, suppose someone argued instead that it should be OK to kill the other person and take their stuff. And were a convicted murderer.
If you’re assuming that they won’t be punished if they convinced the other person, then that’s true. That would be a conflict of interest and hint at them starting with the bottom line.
If you don’t assume that, then it sounds like ad hominem combined with circular logic. Them being a murderer doesn’t mean their argument is wrong. In fact, since they’re living the conclusion, it’s evidence that they actually believe it, and thus that it’s write. Furthermore, them being a murderer is only bad if you already accept the conclusion that it’s not OK to kill the other person and take their stuff.
You can’t say that whenever they are a murderer or not has no relation to the argument they’re making, while you can say that for the face being ugly, though.
I voted your comment up because I agree that the vocabulary is useful for both the person committing the fallacy and (I think this is overlooked) for the person recognizing the fallacy.
However, I think the point of the original quote is probably that when someone points out a fallacy they are probably felling angry and want to insult their interlocutor.
-rekam
That’s not even an example of the ad hominem fallacy.
“You have an ugly face, so you’re wrong” is ad hominem. “You have an ugly face” is not. It’s just a statement. Did the speaker imply the second part? Maybe… but probably not. It was probably just an insulting rejoinder.
Insults, i.e. “Attacking you, not your argument”, is not what ad hominem is. It’s a fallacy, remember? It’s no error in reasoning to call a person ugly. Only when you conclude from this that they are wrong do you commit the fallacy.
So:
A: It’s wrong to stab your neighbor and take their stuff.
B: Your face is ugly.
A: The ugliness of my face has no bearing on moral...
B, interrupting: Didn’t say it does! Your face is still ugly!
They did not logically entail it but they did conversationally implicate it (see CGEL, p. 33 and following, for the difference). As per Grice’s maxim of relation, people don’t normally bring up irrelevant information.
At which point A would be justified in asking, “Why did you bring it up then?” And even if B had (tried to) explicitly cancel the pragmatic implicature (“It’s wrong to stab your neighbor and take their stuff”—”I won’t comment on that; on a totally unrelated note, your face is ugly”), A would still be justified in asking “Why did you change the topic?”
B here is violating Grice’s maxims. That’s the point. He’s not following the cooperative principle. He’s trying to insult A (perhaps because he is frustrated with the conversation). So applying Gricean reasoning to deduce B’s intended meaning is incorrect.
If A asks “why are you changing the subject?”, B’s answer would likely be something along the lines of “And your mother’s face is ugly too!”.
Then he doesn’t get to complain when people mis-get his point.
My point is that Grice’s maxims, useful though they are, do not fully capture how human conversation goes — most notably, those cases in which at least one party has a hostile or uncooperative attitude toward the other. People in such cases do get that they’re being insulted or whatever; A, as you portray him, comes off as simply bad at understanding non-literal meaning (or he is being intentionally obstructive/pedantic).
I contest the empirical claim you are making about human behaviour. That reply in that context very nearly always constitutes arguing against the point the other is making. In particular, the example to which you are replying most definitely is an example of a fallacious ad hominem.
In common practice it does. The rules do change based on attractiveness. (Tangential.)
But A hadn’t specified who the stabber is or who the stabbee is.
The effect of the fallacy can be implied, can’t it?
Can be and usually is (implied).