The Logical-Fallacy Attack
Xrlq has an excellent post on the common rhetorical maneuver of declaring the other side’s argument to be a logical fallacy. This maneuver is just a cheap shot at the other fellow’s emotions designed to shut down debate. And it’s quite effective at that, because the attack is so concise, and formulating the correct response requires a lot of thought and explanation. Now Xrlq has done the thinking and written out the explanation, so the rest of us can benefit from his effort.
Xrlq points out two problems with criticizing an argument as a logical fallacy. The minor point is that it trades unfairly on the different meanings of the word ‘fallacy’. In ordinary usage, it refers to an untrue statement. But in this rhetorical maneuver the speaker is relying on the word’s technical definition within formal logic, which is very different. Even if the listeners understand that the speaker is using ‘fallacy’ in this technical sense, it still has a negative emotional association. The speaker usually understands this, and slip the word in as often as possible which attacking the argument.
Emotional impact aside, the deeper problem with declaring an argument to be a logical fallacy is that it assumes that people can and should think in terms of formal logic. But that’s impossible, as Xrlq explains at some length. The world of formal deductive logic is extremely limited. It’s designed to generate proofs that guarantee the truth of the conclusion, given the truth of the premises. But the real world doesn’t work that way. In the real world, we deal far more with inference and probabilities.
For example, militant gays and their supporters hate the argument that same-sex marriage will lead to the decriminalization and legitimization of other minority sexual practices, such as polygamy. Their favorite attack on that argument is to call it a logical fallacy called the slippery slope.
First, this argument is that it implies a basis in formal logic that doesn’t exist. Within the world of formal symbolic logic, I know of only three true fallacies: denying the antecedent, affirming the consequent, and the fallacy of the undistributed middle. If you look at those three, you’ll get a clear idea of what it means in the formal-logic sense to call something a fallacy. As for the other techniques that are commonly called logical fallacies, such as those listed here, you won’t find them in any books on formal logic. What most people call logical fallacies are really just rhetorical techniques that can easily be abused.
The second problem is that the logical-fallacy attack demands deduction and an extremely high level of certainty where neither is appropriate. The argument that same-sex marriage will lead to polygamy isn’t a guarantee that polygamy can be proven to inevitably occur as a matter of deductive logic. Anything that follows as a matter of deductive logic should be so obvious that no one would waste time discussing it. The SSM-polygamy argument instead calls on the listener to infer that polygamy is significantly more likely given SSM and various other premises that usually go unstated.
So when you’re in a debate, don’t feel the least bit intimidated when someone tells you that your argument is a logical fallacy. See it for the rhetorical cheap shot that it is. The best response is to politely tell your opponent that he doesn’t understand formal logic, and he should perhaps learn some before basing a criticism on it, but you would prefer to discuss the subject at hand. And it’s a pretty safe inferential bet that your opponent in fact knows nearly nothing about formal logic, because few people who have actually studied the topic would so grossly misuse the phrase ‘logical fallacy’.
Ben, I think you are taking things too far. Why isn’t it appropriate to use “logical fallacy” in a colloquial way to indicate the various flaws in arguments outlined on the Wikipedia site? Just because something doesn’t fit the formal logic definition of a logical fallacy doesn’t mean it isn’t a flawed argument. And it hardly seems like a “gross misuse” of the phrase the way it is commonly used.
The dictionary has two definitions for “fallacious”:
1. Containing or based on a fallacy: a fallacious assumption.
2. Tending to mislead; deceptive: fallacious testimony.
Even if we rule out the first on technical grounds, the second still seems legitimate to me.
Comment by Mike S. — May 3, 2006 @ 11:03 am
Mike, the problem isn’t so much with ‘fallacy’, but with ‘logical’. That’s the word that pulls us in the direction of symbolic logic, or at least deductive logic, where neither of those has much of a place in public debate. You could argue that ‘logical’ could also include inductive logic and fuzzy logic, and those would be closer to the form of reasoning that people actually use. But most people don’t understand those types of logic. I certainly don’t.
I agree with you that ‘fallacy’ and ‘fallacious’ still have perfectly acceptable uses in public discourse. I think that ‘fallacious’ tends more towards deception, while ‘fallacy’ better describes a belief that is objectively false. But those uses go to the merits of the point, rather than objecting to its form.
More broadly, the idea that I’m objecting to is that outside of formal logic you can dismiss an argument solely on its form. Ad hominem arguments aren’t always wrong, and neither are appeals to the majority, inferences of causation from correlation, or any of the others. The form neither proves nor disproves the argument.
Comment by BenBateman — May 3, 2006 @ 9:16 pm
More broadly, the idea that I’m objecting to is that outside of formal logic you can dismiss an argument solely on its form. Ad hominem arguments aren’t always wrong, and neither are appeals to the majority, inferences of causation from correlation, or any of the others. The form neither proves nor disproves the argument.
I agree that you can’t dismiss and argument solely on it’s form, but in my experience people tend to use the “logical fallacy” argument in a persuasive sense, not a dispositive one. “You are resorting to ad hominem attacks, which indicate that you don’t have a good response to the argument I put forth”, “Just because two things are correlated doesn’t mean one causes the other”, “Referring to polling data doesn’t tell us what the morally correct position is”, etc.
No doubt some people use the term “logical fallacy” as a form of intimidation, and some people think that a non-formal-logical fallacy is dispositive vis-a-vis their opponents argument, but I think there’s a lot more room for legitimate use of the logical fallacy argument than you indicated in your post.
Comment by Mike S. — May 4, 2006 @ 12:59 pm
Maybe we just have to disagree on this, Mike, but I can think of many situations where the usual logical fallacies flip around: “That’s just an ad hominem attack on Osama Bin Laden, because you don’t have a good response to his reasons for killing all the Jews.”
Or closer to home, we see ad hominem attacks all the time in evolution/ID, running both ways. Both sides say that you should ignore the other side because they’re bad people. And that’s not an inherently flawed argument. Sometimes there really are bad or silly people whose arguments should be ignored, or at least discounted, on that basis. Raelians and Scientologists comes to mind. And those two groups are often used against ID as examples of why ad hominem attacks are appropriate in some circumstances, because science can’t stop to consider every silly theory that some crackpot dreams up. The response to the comparison of ID to the kook isn’t in the argument’s form, but in its substance.
It’s different on correlation and causation, because the attack usually warps what it’s attacking. It’s true that you shouldn’t infer causation solely on correlation, but that doesn’t happen too often. Instead, it’s usually “correlation plus,” meaning that there’s some extra fact that you add to the correlation that raises the likelihood of causation.
For example, height and weight are correlated, and on that basis alone you couldn’t infer any causation. But add some everyday experience, and it seems likely that height causes weight rather than weight causing height, with diet and genetic factors affecting both. So if you’re arguing about causation in height and weight, then I think that it’s unfair to pretend as if people’s beliefs are based solely on correlations, when in fact they’re based on knowledge that is far more extensive but unstated.
You could take it a step further and say that the problem here is in ignoring perfectly valid inductive logic simply because it’s extremely cumbersome and inappropriate in a debate context to trot out all the accumulated personal data points on which each participant bases his inductive conclusions. It may even be impossible for the participants to list those data points, because much of our inductive reasoning is unconscious. We all know intuitively that people can be heavy because they’re tall, but usually aren’t tall because they’re heavy. But how many people can explain the years worth of personal observations on which they base that inference?
Comment by BenBateman — May 4, 2006 @ 2:28 pm
Sounds like I need to add the accusation of mis-using logical fallacies to the list of logical fallacies of which I accuse people.
*grin*
Comment by Bruce — May 7, 2006 @ 11:54 pm