When is ad hom a fallacy?

[GreedyAlgorithm]

JetLeg brings up the question of whether or not learning an arguer was drunk at the time he made his argument should lead us to believe the argument is less likely to be true in this thread (http://forums.randi.org/showthread.php?t=135674). There's something intuitively compelling about the claim that it should, and something else compelling about the claim that it's just an ad hominem.

Here is the difference:

The substance of the argument itself screens off the drunkenness of the arguer. Obviously if the argument is sound then it is sound regardless of the state of the one making the argument. If however we do not know the argument itself, then knowing that the arguer was drunk is very relevant evidence that once we know the argument, we are more likely to find it fallacious. In terms of causality, being drunk is one possible cause of making bad arguments, but is not the cause of any instantiated argument being fallacious. In the causal graph, we'd see

DRUNKENNESS ==> SPECIFIC ARGUMENT MADE ==> FALLACIOUS ARGUMENT

as one subset of the graph. Until we observe the value of SPECIFIC ARGUMENT MADE (i.e. hear the actual argument), if we observe that DRUNKENNESS is higher, we compute that FALLACIOUS ARGUMENT is higher. But if we know the actual value of SPECIFIC ARGUMENT MADE, it screens off the "probability flow" from DRUNKENNESS to FALLACIOUS, and knowing that DRUNKENNESS is higher no longer raises FALLACIOUS.

And so it is ad hom to continue to use evidence of drunkenness as evidence against the argument once you know the argument, but it is not ad hom before. This is why it's fine to decline even reading another of Ray Comfort's arguments - it's so unlikely that the argument is any good, it's a better use of your time to do other things. Did you actually read one, though, the mere fact that it is Ray Comfort's will no longer help you guess whether the argument is good or bad - you know too much now.

(or read this on argument from authority: Argument screens off authority (http://www.overcomingbias.com/2007/12/argument-screen.html))

[Arkayik]

Hic... Who cares...?

On the other hand, if they're drunk enough, will they remember the ad-hom-slur once they sober up?

Still and all, I think 42 is likely to be the best answer you'll get....

Yo, toss another cold one on the Bar-bee-que...!

[nzric]

Bessie Braddock: “Sir, you are drunk.”
Churchill: “Madam, you are ugly. In the morning, I shall be sober.”

[Cavemonster]

It is fallacious to assume that a drunk's statement is likely to be false because it's based on the flawed premise that the majority of statements made by drunks are false.
However, in some areas, this is quite reasonable. These include statements about:
1) Their current ability to drive
2) Whether or not they love you and how much
3) If it really is a good idea to call that ex at 4am
4) We're going to start a band

[linusrichard]

It is fallacious to assume that a drunk's statement is likely to be false because it's based on the flawed premise that the majority of statements made by drunks are false.
Yes, exactly. If you accept the premises, there's nothing wrong with the reasoning that leads to the conclusion. That is, if Jetleg tweaked his premises slightly, his argument would be valid, and not fallacious, but not at all sound.

However, in some areas, this is quite reasonable. These include statements about:
1) Their current ability to drive
2) Whether or not they love you and how much
3) If it really is a good idea to call that ex at 4am
4) We're going to start a band
Dude! A band! Yes! That's going to be so awesome! Oh, man... a band! Totally!

[slingblade]

I read the article.

I do not agree with:

And so it is ad hom to continue to use evidence of drunkenness as evidence against the argument once you know the argument, but it is not ad hom before.

It IS an ad hom, before, during, and after, if you are evaluating the argument in light of the condition of the arguer, and not on the argument's own strengths regardless of maker.

It is even MORE SO an ad hom before, if it prevents you even listening to the argument to evaluate it! For crying out loud, how elementary is that?

It doesn't matter who is making the argument, or what condition they are in. You must judge the argument alone, not the maker.

Whenever you focus on the maker instead of the argument, you commit an ad hominem fallacy. Period. You may be perfectly justified in doing so, but that doesn't make the fallacy disappear!

[linusrichard]

I think it's worth splitting the "argument" up into two parts: the premises, and the reasoning which leads to the conclusion.

If we don't know the truth of the premises, there's nothing wrong with looking at the drunkenness of the proponent as one piece of evidence that will lead us to decide whether we consider them to be true or false. I don't consider that ad hom.

If we don't know the validity of the reasoning (that is, assume the truth of the premises and ask if they lead to the conclusion), that's our own failing. We can't look to the drunkenness of the proponent as having any meaning at all. We have to judge the validity of the argument alone, as I think slingblade was saying. If we don't have the ability to judge the validity of the argument alone, our recourse is to a more capable logician, not to the PBT.

[slingblade]

I think it's worth splitting the "argument" up into two parts: the premises, and the reasoning which leads to the conclusion.

If we don't know the truth of the premises, there's nothing wrong with looking at the drunkenness of the proponent as one piece of evidence that will lead us to decide whether we consider them to be true or false. I don't consider that ad hom.

But that's the very definition of ad hominem!

The argument is not wrong (or right) because Bob is stinking drunk. The argument is wrong because it is wrong.

If a perfectly sober person said the same thing, would it become more true, more correct, more right than if Bob the Stinking Drunk said it?


If we don't know the validity of the reasoning (that is, assume the truth of the premises and ask if they lead to the conclusion), that's our own failing. We can't look to the drunkenness of the proponent as having any meaning at all. We have to judge the validity of the argument alone, as I think slingblade was saying. If we don't have the ability to judge the validity of the argument alone, our recourse is to a more capable logician, not to the PBT.

Exactly.

As I said, you may be emotionally justified in deciding that you don't trust or won't listen to a particular person. But none of that has anything to do with the truth or false values of his statements.

[GreedyAlgorithm]

It is fallacious to assume that a drunk's statement is likely to be false because it's based on the flawed premise that the majority of statements made by drunks are false.
However, in some areas, this is quite reasonable. These include statements about:
1) Their current ability to drive
2) Whether or not they love you and how much
3) If it really is a good idea to call that ex at 4am
4) We're going to start a band
There's a big difference between "is likely to be false" and "is more likely to be false". The first is talking about a total probability, and requires lots of data and evidence to bear out. The second is talking about the direction of change of probability.

[IMST]

You're wrong BECAUSE you're drunk: fallacious ad hom.
You're wrong AND you're drunk: not a fallacious ad hom, though there should be some other reason noted for the wrongness.

[phantomb]

A logical fallacy is by definition an argument that is not logically sound. A logical fallacy is still a logical fallacy even if its conclusion is correct (even if that conclusion is obvious because of another argument).

All logical fallacies are fallacious all of the time.

[GreedyAlgorithm]

As I said, you may be emotionally justified in deciding that you don't trust or won't listen to a particular person. But none of that has anything to do with the truth or false values of his statements.
No, you are logically justified. No weaseling out of it. The reason you're justified is because their arguments are less likely to be right, and you have better things to do. It has everything to do with the truth values of his statements - you have good reason to believe his statements will not contain much truth.

Now suppose he says "Grass is green." Does that make "Grass is green" less likely to be true? Of course not, likelihood is a function of what you know. Once you know the argument, you can evaluate it on its merits. But before you know what was said, you consider it less likely to be true.

How can it possibly be a fallacy to update your beliefs about an unknown statement based on the known speaker when it leads to truer beliefs? That's just silly.

[JoeTheJuggler]

You're wrong BECAUSE you're drunk: fallacious ad hom.
You're wrong AND you're drunk: not a fallacious ad hom, though there should be some other reason noted for the wrongness.
You're saying, I take it, that if the ad hominem remark is not made as part of an argument it can't be a fallacy.

On the other hand, making an ad-hom remark (even if it's simply an insult and not dressed up as an argument) in the context of a debate is still in a way fallacious. It's meant to change the discussion from a debate of ideas into an "argument" (in the non-rhetorical sense) based on emotion and conflicting wills.

So, I'd say anytime an ad hominem comment is made in the context of debate, it is fallacious. Just as it would be fallacious to try to win your debate by knocking your opponent unconscious (what I would call the "might makes right" fallacy, the "my dad can whip your dad" fallacy or the "argumentum ad majorem vim").

NB: I'm just making up this category. There is no conventionally recognized fallacy of this name that I'm aware of.

ETA: Also, if someone said, "You're wrong, and you're drunk" in the context of a debate, I'd take the implied meaning to be, "You're wrong because you're drunk."

[JoeTheJuggler]

A logical fallacy is by definition an argument that is not logically sound. A logical fallacy is still a logical fallacy even if its conclusion is correct (even if that conclusion is obvious because of another argument).

All logical fallacies are fallacious all of the time.

Well said.

The only way I could think of a situation where the statement, "You're drunk" is used properly during an argument would be if the proposition your opponent were arguing is, "I'm sober." Even in this bizarre situation, it would merely be an assertion of the conclusion. If used as an argument, rather than just spelling out what the two positions are, it would be begging the question.

[slingblade]

No, you are logically justified. No weaseling out of it. The reason you're justified is because their arguments are less likely to be right, and you have better things to do. It has everything to do with the truth values of his statements - you have good reason to believe his statements will not contain much truth.

No. That is an emotional response, not a logical one.

To say "I have decided not to waste my time listening to his arguments, because he is less likely to be right" is still an ad hominem, even if that statement is perfectly true. It is still evaluating the man, and not the argument.

Now suppose he says "Grass is green." Does that make "Grass is green" less likely to be true? Of course not, likelihood is a function of what you know. Once you know the argument, you can evaluate it on its merits. But before you know what was said, you consider it less likely to be true.

Which is why you don't dismiss the arguer. You ignore who is making the argument, and focus on the argument itself, because focusing on the person making the argument leads you to exactly this kind of fallacious reasoning.

How can it possibly be a fallacy to update your beliefs about an unknown statement based on the known speaker when it leads to truer beliefs? That's just silly.

How can you "update your beliefs" if you refuse to listen to the speaker because you don't trust him, or feel he is usually wrong? You've screwed yourself out of the chance to learn anything.

Beliefs are often wrong anyway. If you know a thing, you don't have to believe it. If you only believe it, you don't actually know and should be trying to find out.




A logical fallacy is by definition an argument that is not logically sound. A logical fallacy is still a logical fallacy even if its conclusion is correct (even if that conclusion is obvious because of another argument).

All logical fallacies are fallacious all of the time.

Exactly so. Which is why you may emotionally decide a speaker isn't trustworthy, but you are committing a logical fallacy in doing so. Until you evaluate the argument alone, you can say nothing about the argument.

[Piscivore]

So, I'd say anytime an ad hominem comment is made in the context of debate, it is fallacious. Just as it would be fallacious to try to win your debate by knocking your opponent unconscious (what I would call the "might makes right" fallacy, the "my dad can whip your dad" fallacy or the "argumentum ad majorem vim").
That's ad baculum, surely?

Sling, you're brilliant. Couldn't say it better myself.

[Skeptic Ginger]

Here's my example. Last night we were debating Creationists at our Drinking Skeptically Meetup. One of the Creationist gentleman gave his rationale that because 'he' could not explain the evidence for evolution any better than he could support the Biblical version with evidence, it was equally legit that the Bible could be correct.

Obviously the guy was poorly informed about the science involved. While I didn't use that ad hom as an open argument against him, I certainly considered it a valid criticism of his position. His ignorance was the evidence against his viewpoint.

[Piscivore]

Here's my example. Last night we were debating Creationists at our Drinking Skeptically Meetup. One of the Creationist gentleman gave his rationale that because 'he' could not explain the evidence for evolution any better than he could support the Biblical version with evidence, it was equally legit that the Bible could be correct.

That's argument from ignorance, already a fallacy.

Obviously the guy was poorly informed about the science involved. While I didn't use that ad hom as an open argument against him, I certainly considered it a valid criticism of his position. His ignorance was the evidence against his viewpoint.
It would be a fallacy no matter who said it, therefore it is not ad hom.

[jmcvann]

Did Einstein become stupid when he drank?

[JoeTheJuggler]

That's ad baculum, surely?

As they say, a derringer beats a royal flush. . . .

I thought I was making something up!

Sub sole nihil novi est.

[Skeptic Ginger]

That's argument from ignorance, already a fallacy.


It would be a fallacy no matter who said it, therefore it is not ad hom.Yes but if I expand it and say he didn't know much about biology so I'm certainly not going to buy what he's selling about physics, it becomes and ad hom, doesn't it?

Perhaps the philosophical level of this discussion is over my head.

[linusrichard]

But that's the very definition of ad hominem!
I don't think it is. But maybe I was unclear - I'm not talking about the argument - I'm talking about the premises.

The argument is not wrong (or right) because Bob is stinking drunk. The argument is wrong because it is wrong.
Right, but I'm talking about the premises here. And obviously the drunkenness of the proponent doesn't affect the truth or falsity of the premise, but - if we don't know the truth or falsity of the premise, we may consider the credibility of the proponent in deciding whether or not to accept it as true or false. And when considering the credibility of the proponent, we can look at the drunkenness of the proponent as one piece of evidence that goes to credibility.

Let me come up with some clumsy examples:
Alan walks up to you, drunk. You have no idea where he lives. "Can I take you home, Alan?" "Sure, take me to 123 Aspen." I would take him to 123 Aspen.
Bob walks up to you, stone sober. You are pretty sure he lives at 123 Birch. "Can I take you home, Bob?" "Sure, take me to 132 Birch." I would take him to 132 Birch.
Carl walks up to you, drunk. You are pretty sure he lives at 123 Cedar. "Can I take you home Carl?" "Sure, take me to 132 Cedar." I would probably take him to 123 Cedar.

Now, these don't look like arguments, which is why my examples are clumsy, but let's turn them into arguments:
1. Major premise: You should take me to the address which is my home.
2. Minor premise: The address which is my home is ______.
3. Conclusion: You should take me to _____.

Now, all three arguments are perfectly valid, and their validity can be judged without reference to the men's drunkenness. But we see that when we evaluate each man's minor premise, we can take the drunkenness into account. In Alan's case, we have no better alternative, so even though he is drunk, the best we can do is assume his minor premise is correct. In Bob's case, we have a reason to think he is wrong (our previous idea about his address), but a better reason to think he is right (a sober person is unlikely to misstate his address). And in Carl's case, we have a reason to think he's right (a person knows his address), but a better reason to think he's wrong (our previous idea about his address combined with the possibility that his drunkenness has caused him to transpose numbers).

If a perfectly sober person said the same thing, would it become more true, more correct, more right than if Bob the Stinking Drunk said it?
Of course not. But that's not what I'm saying. What I'm saying is that we can consider drunkenness in evaluating the probability that the conclusion is correct, not (obviously) because it is relevant to the validity of the reasoning, but because it might be relevant to whether we trust the premises. Just as Bob and Carl presented us with identical situations, other than the drunkenness, but it was reasonable to reject Carl's conclusion and accept Bob's, based on that difference.

[slingblade]

Right, but I'm talking about the premises here. And obviously the drunkenness of the proponent doesn't affect the truth or falsity of the premise, but - if we don't know the truth or falsity of the premise, we may consider the credibility of the proponent in deciding whether or not to accept it as true or false. And when considering the credibility of the proponent, we can look at the drunkenness of the proponent as one piece of evidence that goes to credibility.

Okay, I'll take a different tack.

There never comes a time when you weigh an argument by how much you trust or distrust the person making it.

Who makes the argument has no bearing on the truth or falsity of either the premises or the conclusion.

Validity refers only to the form of the argument. An argument can be valid, even when the premises are false.

Soundness refers only to content. An argument is sound if its premises are true. But it can still be invalid, if true premises lead to a false conclusion.

Any time you step out of that frame and postulate that an argument is either invalid, unsound, or both, because of who is making the argument, you have committed an ad hominem fallacy.

I note that all the examples given for this are not actually arguments, but statements placed in argumentative form. I'd opine this is because it's difficult if not impossible to come up with an actual argument in which the condition, reputation, or identity of the arguer matters to the soundness and validity of the argument.

One more time: trusting the arguer is fallacious because you do not know who you can or cannot trust. Experts make mistakes; they get it wrong; they misspeak. The opposite of ad hominem is the appeal to authority: "I trust what Dr. Right says, because he's an expert in his field," is just as fallacious as "I don't trust Rush Limbaugh; he's a bigot."

If you resort to these fallacies out of your own ignorance, it doesn't make them any less fallacious.

Mr. X is arguing about quantum physics.
I know nothing about quantum physics.
Mr. X is drunk.
I know drunks are unreliable thinkers.

Therefore, I will disregard what Mr. X says, because you can't trust a drunk...



...is always an ad hominem fallacy.

[Piscivore]

Yes but if I expand it and say he didn't know much about biology so I'm certainly not going to buy what he's selling about physics, it becomes and ad hom, doesn't it?
Yeah, if you elect to disregard his arguments because of who he is (or who you think he is), not on the content or quality of the arguments themselves, that's ad hom.

Perhaps the philosophical level of this discussion is over my head.
I doubt it, it's basic logic.

[arthwollipot]

An ad hominem is a logical fallacy. A mere insult is not.

There is a difference between an ad hominem and an insult.

Actually, I think someone has already pointed this out.

Carry on!

[linusrichard]

Okay, I'll take a different tack.

There never comes a time when you weigh an argument by how much you trust or distrust the person making it.
When you say "argument" here, do you refer to the reasoning leading from the premises to the conclusion (validity), or do you refer to the premises, the reasoning, and the conclusion, taken together (soundness)?
Who makes the argument has no bearing on the truth or falsity of either the premises or the conclusion.

Validity refers only to the form of the argument. An argument can be valid, even when the premises are false.
I agree with both of these.
Soundness refers only to content. An argument is sound if its premises are true. But it can still be invalid, if true premises lead to a false conclusion.
Sorry - this is kind of a minor quibble, but this is a mistake. An argument is sound if its premises are true and it is valid. A sound argument is always valid. An invalid argument cannot be sound. But it's not really an important point for this discussion.

Any time you step out of that frame and postulate that an argument is either invalid, unsound, or both, because of who is making the argument, you have committed an ad hominem fallacy.
Yes.

I note that all the examples given for this are not actually arguments, but statements placed in argumentative form. I'd opine this is because it's difficult if not impossible to come up with an actual argument in which the condition, reputation, or identity of the arguer matters to the soundness and validity of the argument.
Well, I'd say that I turned them into arguments, but I won't argue the point. You are absolutely right that the condition, reputation, or identity of the arguer never matters as to the soundness and validity of the argument. But you can't always look at an argument and tell whether or not it is sound. (You can, if you are smart enough, and I think we both are, look at an argument and tell whether or not it is valid.) So your decision whether or not to believe an argument is sound is actually not dependent on whether the argument actually is sound, but based on what little information you have indicating its soundness. And the reliability of the proponent of a premise is one of those pieces of information.

One more time: trusting the arguer is fallacious because you do not know who you can or cannot trust. Experts make mistakes; they get it wrong; they misspeak. The opposite of ad hominem is the appeal to authority: "I trust what Dr. Right says, because he's an expert in his field," is just as fallacious as "I don't trust Rush Limbaugh; he's a bigot."
"Just as fallacious" in the sense that neither is fallacious. "Dr. Right is correct because he's an expert in his field" is fallacious. "Rush Limbaugh is wrong because he's a bigot" is fallacious. But using reputation to decide whom to trust is not fallacious; in fact, it is perfectly reasonable, although it may, as you note, lead you to believe some things that turn out not to be true.

If you resort to these fallacies out of your own ignorance, it doesn't make them any less fallacious.

Mr. X is arguing about quantum physics.
I know nothing about quantum physics.
Mr. X is drunk.
I know drunks are unreliable thinkers.

Therefore, I will disregard what Mr. X says, because you can't trust a drunk...

...is always an ad hominem fallacy.

I disagree. If you say "Mr. X is wrong, because drunks are unreliable thinkers," then you've committed an ad hominem fallacy. If you say "I will disregard what Mr. X says, because you can't trust a drunk," you've made a choice (arguably a reasonable choice) about what to believe given imperfect information.

Unless you're going to remain completely agnostic about any fact which you can't personally confirm, and, by extension, any argument that has such a fact as a premise, you're going to have to make choices about what to believe. Some of those choices will be correct, and some will not, and there is no fallacy in using the proponent's mental condition or reputation as one of your tools in trying to make your choice.

It is true that when Bob the drunk makes an argument, that argument Really Is Sound or Unsound completely regardless of his state of insobriety. But if you don't have access to that information, there's nothing wrong with betting on unsound.

[UnrepentantSinner]

You're a drunk moron and your argument sucks because you're a drunk moron. ad hom
You're a drunk moron and your agrument sucks for reasons A, B, C, C1, C2, D and E. not ad hom

[ImaginalDisc]

You're saying, I take it, that if the ad hominem remark is not made as part of an argument it can't be a fallacy.

On the other hand, making an ad-hom remark (even if it's simply an insult and not dressed up as an argument) in the context of a debate is still in a way fallacious. It's meant to change the discussion from a debate of ideas into an "argument" (in the non-rhetorical sense) based on emotion and conflicting wills.

So, I'd say anytime an ad hominem comment is made in the context of debate, it is fallacious. Just as it would be fallacious to try to win your debate by knocking your opponent unconscious (what I would call the "might makes right" fallacy, the "my dad can whip your dad" fallacy or the "argumentum ad majorem vim").

NB: I'm just making up this category. There is no conventionally recognized fallacy of this name that I'm aware of.

ETA: Also, if someone said, "You're wrong, and you're drunk" in the context of a debate, I'd take the implied meaning to be, "You're wrong because you're drunk."

There's no such thing as an ad hominem comment. Logical fallacies are fallacies of reasoning, not commentary.

"Bob, you are wrong because you are stupid."

"Bob's opinion is worthless, because he's stupid."

"Bob's infamous for his stupidity, and now he brings up this argument? Well, we all know what's that worth."

From explicit to implicit, those are arguments, and therefore ad hominems.

"Bob is stupid. He is also wrong."

That's not an ad hominem.

"Bob, you're a jackass, a waste of oxygen, and an embarrassment to humanity."

That's just a plain ol' insult.

"Bob, the moon landings were not faked. The 'strange' shadows are a behavior of shadows in perspective familiar to any photographer. The reason the stars aren't visible in the photos is because they're quite dim compared to the brutal glare of the sun. Also, there are mirrors the astronauts placed there that are being used today to measure its distance.

Jeez, Bob. You're stupid."

That's perhaps unkind, but not an ad hominem.

Lastly, a caveat: Just because an argument is fallacious doesn't mean it's false, just badly supported.

[slingblade]

Sorry - this is kind of a minor quibble, but this is a mistake. An argument is sound if its premises are true and it is valid. A sound argument is always valid. An invalid argument cannot be sound. But it's not really an important point for this discussion.


Gah, you're absolutely right. I did think I had it wrong somewhere, but convinced myself I was good to go, anyway. Thanks for the correction. :)


Well, I'd say that I turned them into arguments, but I won't argue the point. You are absolutely right that the condition, reputation, or identity of the arguer never matters as to the soundness and validity of the argument. But you can't always look at an argument and tell whether or not it is sound. (You can, if you are smart enough, and I think we both are, look at an argument and tell whether or not it is valid.) So your decision whether or not to believe an argument is sound is actually not dependent on whether the argument actually is sound, but based on what little information you have indicating its soundness. And the reliability of the proponent of a premise is one of those pieces of information.

I know it is, regarding your last remark, but it is risky and often misguided thinking to do so, which is why it's a fallacy.

Again, you aren't supposed, logically speaking, to believe or disbelieve arguments. You're supposed to prove them, to use logical proofs, truth tables, and formal examination (modus ponens, modus tollens, etc.).


"Just as fallacious" in the sense that neither is fallacious. "Dr. Right is correct because he's an expert in his field" is fallacious. "Rush Limbaugh is wrong because he's a bigot" is fallacious. But using reputation to decide whom to trust is not fallacious; in fact, it is perfectly reasonable, although it may, as you note, lead you to believe some things that turn out not to be true.

I think the hairs are being split too finely, here. It ends up being the same result.

The whole idea of the fallacy of ad hominem is to keep you from belief and lead you to knowledge. You aren't being logical to believe someone is honest, accurate, what have you, based on reputation.

I keep trying to get you to understand that while what you suggest in your entire argument is accurate from an emotional standpoint, and is indeed what many people do most of the time, it is still a logical fallacy.

So what's the difference, in terms of results, for you to say "I don't tend to trust Rush Limbaugh, because he's shown evidence of being a bigot," and "Rush Limbaugh is wrong because he's a bigot?"

Don't both lead you to not listening to the man's individual arguments, but rather dismissing them and him because you "don't trust him?"

Fine, don't trust him, but examine his arguments anyway! Ad hominem is simply a warning to you that you're letting your own bigotry get in the way of your logic!

I disagree. If you say "Mr. X is wrong, because drunks are unreliable thinkers," then you've committed an ad hominem fallacy. If you say "I will disregard what Mr. X says, because you can't trust a drunk," you've made a choice (arguably a reasonable choice) about what to believe given imperfect information.

It's the same thing! Unless you're telling me you knowingly disregard correct information as often as you do incorrect?

Unless you're going to remain completely agnostic about any fact which you can't personally confirm, and, by extension, any argument that has such a fact as a premise, you're going to have to make choices about what to believe. Some of those choices will be correct, and some will not, and there is no fallacy in using the proponent's mental condition or reputation as one of your tools in trying to make your choice.

It is true that when Bob the drunk makes an argument, that argument Really Is Sound or Unsound completely regardless of his state of insobriety. But if you don't have access to that information, there's nothing wrong with betting on unsound.

It doesn't make an ad hom non-fallacious. But we're just going to have to disagree. We're talking at slightly cross-purposes here.

I'm trying to tell you that even though we all do what you suggest, it's still a fallacy to do so. That you have, or seem to have no choice doesn't make it less fallacious, but may make it more humanly, emotionally understandable.

Logic doesn't care that you don't have the time or motivation or inclination to check out every person's argument without resorting to an ad hom fallacy. Your inability doesn't render the logic moot.

[linusrichard]

@slingblade: you deserve a better response than I have the time (and maybe the ability) to give. I guess we can agree to disagree for now.

I guess what I would say in the short time I will allow myself is that what we're talking about is not a logical fallacy because it's not a logical anything. You say it's correct from an emotional standpoint, but I say the emotional standpoint is the only relevant standpoint to look at it from, which means it's correct from the only relevant standpoint. It's the difference between saying "I'm going to use Bob's drunkenness (or Rush's bigotry, or Dr. Right's c.v.) to evaluate his argument" (fallacious), and "I'm going to use Bob's drunkenness (etc.) because I'm unable to evaluate his argument." Using "evaluate" in kind of a strict sense. Does this help? It's all I can do at the moment...

[arthwollipot]

You're a drunk moron and your argument sucks because you're a drunk moron. ad hom
You're a drunk moron and your agrument sucks for reasons A, B, C, C1, C2, D and E. not ad homThis is what I was trying to say, earlier.

The first is an ad-homimen, and a logical fallacy.
The second is an insult, coupled with a logially valid argument.

Not all insults are ad-homs. It could be argued that all ad-homs are insults though.

[linusrichard]

It could be argued that all ad-homs are insults though.

If someone as smart as you says so, it must be true.



;)

[westprog]

Again, you aren't supposed, logically speaking, to believe or disbelieve arguments. You're supposed to prove them, to use logical proofs, truth tables, and formal examination (modus ponens, modus tollens, etc.).



Belief and disbelief are not contrary to logic - they simply operate in a different sphere. Logic can inform belief, but belief can't be logically unsound.


I think the hairs are being split too finely, here. It ends up being the same result.

The whole idea of the fallacy of ad hominem is to keep you from belief and lead you to knowledge. You aren't being logical to believe someone is honest, accurate, what have you, based on reputation.


Evaluating someone's honesty is something one does every day - and it can be based on logic, but cannot wholly rely on it.

It isn't contrary to logic to make predictions of future behaviour based on what happened in the past. That's the entire basis of science.



I keep trying to get you to understand that while what you suggest in your entire argument is accurate from an emotional standpoint, and is indeed what many people do most of the time, it is still a logical fallacy.


What people do can't be a logical fallacy. Making a choice about whether or not to address a particular argument is quite a different matter, logically speaking, to claiming that the argument is false.

So what's the difference, in terms of results, for you to say "I don't tend to trust Rush Limbaugh, because he's shown evidence of being a bigot," and "Rush Limbaugh is wrong because he's a bigot?"

Don't both lead you to not listening to the man's individual arguments, but rather dismissing them and him because you "don't trust him?"

Fine, don't trust him, but examine his arguments anyway! Ad hominem is simply a warning to you that you're letting your own bigotry get in the way of your logic!



But it isn't possible to examine everybody's arguments!

For example - it's been claimed that Obama was sworn in on a Koran. Am I to give that claim the same attention as a CNN report that Obama and the Chief Justice muddled the oath and had to repeat it the next day? I don't think my choice of accepting one and not accepting the other contention is an example of "bigotry" or shutting my mind. I certainly don't think it's contrary to logic. Logic doesn't enter into it.

It's the same thing! Unless you're telling me you knowingly disregard correct information as often as you do incorrect?



Nearly all decisions are made from imperfect information. That's the nature of things. Making a judgement on incomplete grounds is not contrary to logic.

It doesn't make an ad hom non-fallacious. But we're just going to have to disagree. We're talking at slightly cross-purposes here.

I'm trying to tell you that even though we all do what you suggest, it's still a fallacy to do so. That you have, or seem to have no choice doesn't make it less fallacious, but may make it more humanly, emotionally understandable.


Doing something has nothing to do with logic, and cannot be a fallacy. Fallacy's exist only within the very enclosed world of logic. They do not exist in the world of human action.


Logic doesn't care that you don't have the time or motivation or inclination to check out every person's argument without resorting to an ad hom fallacy. Your inability doesn't render the logic moot.

Logic doesn't care what you do. Logic is its own enclosed system.

A logical judgement can be logically wrong. A real life belief cannot. A real life belief may be based on a logical fallacy - for example, a claim that drunks never tell the truth would clearly be logically flawed - but the flaw lies in the logic, not in the action taken.

[JoeTheJuggler]

There's no such thing as an ad hominem comment. Logical fallacies are fallacies of reasoning, not commentary.
Ad hominem is Latin for "to the man". So yes, any personal comment directed at the person (as opposed to the topic) may be called an "ad hominem" remark. Now whether that is the fallacy of argumentum ad hominem is another question.

In a formal debate, any ad hominem remark can be see as fallacious if it's irrelevant to the question. If it's not presented as part of an argument then it is irrelevant. In a way, most informal fallacies can be boiled down to the fact that they're irrelevant.

In most kinds of "debate" things aren't so formalized.

If there was, for example, a protest outside the White House against Bush's decision to invade and occupy Iraq, and someone held a sign that said, "Bush is a Moron", would you consider that sign to be an example of the ad hominem fallacy?

I say it is, because it's implied to be an argument as to why his decision was wrong.

If the situation is somewhere in between, like say a discussion on the JREF forum, the difference between saying, "You're drunk and you're wrong" and saying "You're wrong because you're drunk" is insignificant.

I don't think you can claim an ad hominem remark is not park of the argument (in the context of an argument) merely because you leave out the word "because".

In fact, even in legitimate arguments, those sorts of connections between propositions are frequently left out in informal speech (or forum postings).

[Dymanic]

Belief and disbelief are not contrary to logic - they simply operate in a different sphere.Suppose the proposition is: "Biological organisms exhibit irreducible complexity", and a supporting argument is: "This credentialled biochemist says so, and offers extensive evidence". An argument from authority -- but a valid one, as the authority is recognized as an expert, and is speaking on matters within his field. This seems to be an invitation to an ad hominem refutation: "Yes, but the core issues turn on interpretations of evidence, and the expert's interpretations are biased by his religious beliefs".

[Skeptic Ginger]

Yeah, if you elect to disregard his arguments because of who he is (or who you think he is), not on the content or quality of the arguments themselves, that's ad hom....Yes, but in this hypothetical situation, he brings up something I am less familiar with the specifics of, the second law of thermodynamics. I get it enough to understand the concept but not enough to argue the specifics.

But I know the guy's science cred is very low. I don't need to bother looking into the argument he is making. I've heard these 'believers' before. They have no cred.

At some point one recognizes the pattern, whether it be simple woo embracing, CT mania, or "I got God" witnessing. You draw conclusions on the argument based on the pattern you see in the arguer.

[Skeptic Ginger]

Suppose the proposition is: "Biological organisms exhibit irreducible complexity", and a supporting argument is: "This credentialled biochemist says so, and offers extensive evidence". An argument from authority -- but a valid one, as the authority is recognized as an expert, and is speaking on matters within his field. This seems to be an invitation to an ad hominem refutation: "Yes, but the core issues turn on interpretations of evidence, and the expert's interpretations are biased by his religious beliefs".It's hard to go with your example knowing the science does not support the credentialed researcher's conclusions.

However, we can easily be fooled into accepting arguments, not from authority, but from trust in a person's expertise without verification.

An argument from authority would be lacking any research. Trust in their expertise would be accepting their research as valid without verification.

[Silentknight]

This ought to be a fairly simple issue. An ad hominem attack is always a fallacy, because the arguer's personal life and conditions have no logical bearing on the validity of his/her arguments. If you're arguing from logic, then attacking the person doesn't demonstrate anything relevant to the argument. If you're arguing from credibility, then the game might change a little, but it still boils down to whether the argument can stand on its own merit. A drunk person can be right, and a sober person can be wrong.


I've never been drunk myself, mind you, but doesn't drunkenness only make you more likely to say things without filtering them first? I wasn't aware that intoxication automatically turned all your statements into lies. It can alter your perception and make you less likely to report on events accurately, but that's not the same as outright lying.

[Dymanic]

It's hard to go with your example knowing the science does not support the credentialed researcher's conclusions."The science" can often support more than one conclusion, and credentialed experts differ on conclusions all the time. That's not the issue. Both the argument from authority and the ad hominem argument take place at a level above that of actual research. Ideally, we'd all be equally experts, equally in possession of all the data, and none of us would ever need to defer to anyone else's opinion.

In reality, even experts rely on one another to an ever-increasing degree. So it really does become a matter of credibility. It's also a matter of expedience. So-and-so's support of a position can be offered as an argument without the need to unpack every detail of his work, so long as it is agreed that he is an expert in the relevant field. If such an argument from authority can be valid, why is it not equally valid to refute it by questioning either the expertise of that authority or his ability to form unbiased conclusions?

[Beerina]

Bessie Braddock: “Sir, you are drunk.”
Churchill: “Madam, you are ugly. In the morning, I shall be sober.”

Ahhh, good ol' Winnie.



Lady Nancy Astor: Winston, if you were my husband, I'd put arsenic in your morning coffee.

Churchill: Madam, if you were my wife, I'd drink it.

[JetLeg]

@slingblade: you deserve a better response than I have the time (and maybe the ability) to give. I guess we can agree to disagree for now.


I'd be very glad to see your whole reply.

[JetLeg]

JetLeg brings up the question of whether or not learning an arguer was drunk at the time he made his argument should lead us to believe the argument is less likely to be true in this thread (http://forums.randi.org/showthread.php?t=135674). There's something intuitively compelling about the claim that it should, and something else compelling about the claim that it's just an ad hominem.

Here is the difference:

The substance of the argument itself screens off the drunkenness of the arguer. Obviously if the argument is sound then it is sound regardless of the state of the one making the argument. If however we do not know the argument itself, then knowing that the arguer was drunk is very relevant evidence that once we know the argument, we are more likely to find it fallacious. In terms of causality, being drunk is one possible cause of making bad arguments, but is not the cause of any instantiated argument being fallacious. In the causal graph, we'd see

DRUNKENNESS ==> SPECIFIC ARGUMENT MADE ==> FALLACIOUS ARGUMENT

as one subset of the graph. Until we observe the value of SPECIFIC ARGUMENT MADE (i.e. hear the actual argument), if we observe that DRUNKENNESS is higher, we compute that FALLACIOUS ARGUMENT is higher. But if we know the actual value of SPECIFIC ARGUMENT MADE, it screens off the "probability flow" from DRUNKENNESS to FALLACIOUS, and knowing that DRUNKENNESS is higher no longer raises FALLACIOUS.

And so it is ad hom to continue to use evidence of drunkenness as evidence against the argument once you know the argument, but it is not ad hom before. This is why it's fine to decline even reading another of Ray Comfort's arguments - it's so unlikely that the argument is any good, it's a better use of your time to do other things. Did you actually read one, though, the mere fact that it is Ray Comfort's will no longer help you guess whether the argument is good or bad - you know too much now.

(or read this on argument from authority: Argument screens off authority (http://www.overcomingbias.com/2007/12/argument-screen.html))

P is the statement "The argument of the drunk is probably fallacious".

According to your claim, before you know the actual argument, P is true. But suddenly, when you know it, then suddenly you have to forget that P is true? It is absurd - if you know P is true, you cannot claim suddenlly you don't know P to be true.

[slingblade]

Ignore all the arguing drunks you like, JetLeg. I promise, you won't miss a thing by doing so, and absolutely no one will mind.

[linusrichard]

Ahhh, good ol' Winnie.



Lady Nancy Astor: Winston, if you were my husband, I'd put arsenic in your morning coffee.

Churchill: Madam, if you were my wife, I'd drink it.

Dave Barry (IIRC) had a couple good parodies of this.

"Madam, if you were my wife, I'd hit you in the face."

"Madam, I may be drunk, but *bleargghhh*" all over her dress.

[JetLeg]

I don't think it is. But maybe I was unclear - I'm not talking about the argument - I'm talking about the premises.


Right, but I'm talking about the premises here. And obviously the drunkenness of the proponent doesn't affect the truth or falsity of the premise, but - if we don't know the truth or falsity of the premise, we may consider the credibility of the proponent in deciding whether or not to accept it as true or false. And when considering the credibility of the proponent, we can look at the drunkenness of the proponent as one piece of evidence that goes to credibility.

Let me come up with some clumsy examples:
Alan walks up to you, drunk. You have no idea where he lives. "Can I take you home, Alan?" "Sure, take me to 123 Aspen." I would take him to 123 Aspen.
Bob walks up to you, stone sober. You are pretty sure he lives at 123 Birch. "Can I take you home, Bob?" "Sure, take me to 132 Birch." I would take him to 132 Birch.
Carl walks up to you, drunk. You are pretty sure he lives at 123 Cedar. "Can I take you home Carl?" "Sure, take me to 132 Cedar." I would probably take him to 123 Cedar.

Now, these don't look like arguments, which is why my examples are clumsy, but let's turn them into arguments:
1. Major premise: You should take me to the address which is my home.
2. Minor premise: The address which is my home is ______.
3. Conclusion: You should take me to _____.

Now, all three arguments are perfectly valid, and their validity can be judged without reference to the men's drunkenness. But we see that when we evaluate each man's minor premise, we can take the drunkenness into account. In Alan's case, we have no better alternative, so even though he is drunk, the best we can do is assume his minor premise is correct. In Bob's case, we have a reason to think he is wrong (our previous idea about his address), but a better reason to think he is right (a sober person is unlikely to misstate his address). And in Carl's case, we have a reason to think he's right (a person knows his address), but a better reason to think he's wrong (our previous idea about his address combined with the possibility that his drunkenness has caused him to transpose numbers).

...

Of course not. But that's not what I'm saying. What I'm saying is that we can consider drunkenness in evaluating the probability that the conclusion is correct, not (obviously) because it is relevant to the validity of the reasoning, but because it might be relevant to whether we trust the premises. Just as Bob and Carl presented us with identical situations, other than the drunkenness, but it was reasonable to reject Carl's conclusion and accept Bob's, based on that difference.


Good post. I agree with that.

But even with regards to the validity of the drunk's logic - there is a larger probability that his arguments will be invalid.


I will give a better example than a drunk - the URL
http://jmm.aaa.net.au/articles/13209.htm

The arguments there are bad ones. If you read 5 arguments from there, are you justified in becoming tired and coming to the conclusion that the rest of the arguments will have bad logic as well? Yes. (I think it just analogous to my example of drawing lots of black balls from a box, and concluding that probably the next one will be also black. )

What I said above is common sense, I think. If a certain person is famous for making arguments with flawed logic all the time, it is probable that his next argument will also have bad logic.

The step that is more problematic is to take a specific argument that he makes and to claim that since it is an argument by such a person, there is a high probability that this argument is false.

[JetLeg]

Ignore all the arguing drunks you like, JetLeg. I promise, you won't miss a thing by doing so, and absolutely no one will mind.

Slingblade, the answer to my question is not that trivial, so you do not do justice to me, dismissing derogatingly.

[JetLeg]

Greedy Algorythm actually made a really good point, and I think it resolves the question.

The argument

The vast majority of X's arguments are invalid
P is an argument by X
---
It is probable that P is invalid.

Is a valid inductive argument, and not an ad hom. However, when one actually judges the specific argument P, since it is possible that it is valid, then one has to judge by its own merit, and not by association with X. It is a bit complicated to put in words, but can be easily explained using the analogy of black and white balls :


If there is a box with 99% black balls, there is a 99% probability that a randomly taken out ball will be black. However, when you take a ball, and try to determine whether it is a black ball or not, you use your sight - you do nto use the probability. You cannot take out a ball, and not judge whether it is black or white "on its own merit". You cannot take a white ball, look at it and declare "there is 99% probability that this ball is black"!!

And linusrichard has a sound point as well, with regards to the premises.