I still remember the thrill of first encountering a summary of Carl Sagan's Baloney Detection Kit. Reading through the list of logical fallacies, I could feel a change come over my being and my posture: my biceps bulged, my abs hardened into a carapace, and my gonads turned to solid granite. I had discovered the secret weapons cache of the elite commandoes of reason, and now I felt invincible. Armed with Sagan's checklist, I was sure I could survive any argument undefeated. Creationists, paranormalists and fanboys alike would fall before my big guns of logic: I'd mow them all down like Arnie in Terminator 2.

But when I went on Internet forums and saw the Baloney Detection Kit in action, I was shocked and puzzled. Thousands of other people had discovered the secret weapons cache, but they were not, as I had expected, blowing away all before them. In their hands, the logical fallacy terms did not look like decisive weapons; instead, they looked clumsy, cumbersome, easy to outmanoeuvre. Why did people who deployed these terms always look so rigid, so predictable, so feeble? Why did people who avoided them look so confident in comparison, so much more in command of their resources, so much more mature? Their arguments seemed to possess an inner strength; the baloney detectors, by contrast, only had strength in numbers. Using Sagan's checklist, I thought I'd have the gods of reason and logic on my side; instead, I just had a bunch of bullying creeps.

I became rapidly disillusioned with the power of logical fallacies, and over time, my disillusionment has only grown. These days, I'm convinced that they have no place in most arguments, and in the rest of this article I'll explain why.

LOGICAL FALLACIES ARE USUALLY IRRELEVANT OR CITED INCORRECTLY

As both my regular readers will know, I made my first million dollars by writing a webpage explaining why the term ad hominem is so often incorrectly used. But I could have written the same for any other fallacy in the Baloney Detection Kit; ad hominem is only the most frequently invoked. Even on the rationalwiki page about these logical fallacies, several of the illustrative examples are incorrect. It's very unusual to see a logical fallacy identified correctly in the wild, and it's rarer still to see one identified to any good purpose.

Logical fallacies are only relevant in certain narrow rhetorical modes and contexts. Specifically, logical fallacies are only relevant in rational arguments. If a speaker is trying to convince you of the gut-wrenching pathos of a situation, then it's irrelevant to call them out on an Appeal to the Emotion, because appealing to the emotion is exactly what they're trying to do. And even if a speaker is trying to convince you of the mind-wrenching logos of a situation, most logical fallacies only come into play when the speaker's language is intended to be taken literally. If a speaker is employing irony, hyperbole, understatement, whimsy, counterfactual conjecture, or any other of the wonders of figurative language that defy semantic nit-picking, then your Baloney Detection Kit has nothing to contribute. To further complicate matters, rhetorical modes and contexts can shift within an argument. At times a speaker can be using literal language, at other times figurative; at times they can appeal to the emotions, at other times reason, at other times both, at other times neither; at times they may not even be arguing at all. Following the shifting contexts of an argument requires careful attention and a note of empathy. Baloney detectors who indiscriminately lob in their logical fallacy terms tend to look like robots who don't understand human speech.

Humans typically communicate in a way that resists shallow logical analysis. In a real conversation, people use words rather than terms, make utterances rather than sentences, and employ a wider variety of inference methods than modus ponens. A great deal of what is communicated and inferred in a conversation depends on context; the speakers and audience, their history, their shared knowledge and confidences, the feelers they lay out to establish mutual trust and rapport. Poking into this with your ad hominem stick betrays an ignorance of the way people actually communicate, and ignorance in general. The most sedulous baloney detectors are guilty of the same literalism as the bible-thumpers they ridicule.

CITING LOGICAL FALLACIES IS USUALLY COWARDLY AND LAZY

In coming up with the Baloney Detection Kit, Carl Sagan wanted to encourage critical thinking; however, in using the kit, it seems most baloney detectors want to reject it. Far too often, logical fallacies are invoked in order to run away from an argument. Rather than engage with the ideas in a text, it's much easier to skim through it trying to spot a quick fallacy; and once a fallacy is found, a baloney detector can safely ignore everything else.

I must admit that I have often been on the receiving end of this tendency. Any time one of my webpages makes it to a discussion board, I see people quote a line and say "Ad Hominem!" or "Begging the Question!" or "Faulty Premise!" and say no more; these people seem to have no other opinions on the text. And then there are people on Reddit who won't read my Ender's Game review because someone told them there's a contradiction somewhere. I say there is "no dramatic tension" at one point in the review, and " [the writer] makes the pages turn" at another; and this apparent contradiction presumably negates the whole argument.

As it happens, I don't think there is a contradiction in my Ender's Game review — dramatic tension isn't the only thing that makes pages turn — but even if there were, its arguments probably wouldn't topple down. This is because an argument is not the same thing as a mathematical proof, and arguing is not the same thing as proof by contradiction. In the real world, it's not sufficient simply to identify a fallacy in an argument. You've also got to think why the fallacy is a problem in that particular instance, and what consequences it could have for the rest of the argument; often it will have less consequence than you might think. If you think you've dismissed an opposing argument just because you think you've seen a fallacy, then you're deluding yourself. But then, lot of baloney detectors seem to prefer self-delusion to critical thought — and there's an irony I've laboured already.

LOGICAL FALLACIES ARE OFTEN USED TO EXCLUDE

The most troubling aspect of logical fallacies is their use in suppressing uncomfortable ideas and viewpoints, and this can happen whether they are invoked correctly or not. I've seen countless examples of fallacies being called upon to dismiss other people's opinions and ride over their emotions. Used in this way, they are tools of power, summoned to establish and protect a self-serving clique.

By way of illustration, suppose you are a guy who identifies as a skeptic. And then suppose you encounter a woman who tells you that because of the insults she has received from guys in the skeptic community, she has decided that the skeptic movement is fundamentally sexist. Here are a dozen ways not to respond:

"You're cherry-picking examples of the worst behaviour. Most of us are nice guys."

"You're concocting a straw man of skepticism. Most of us are nice guys."

"I think you're guilty of confirmation bias here — concentrating on all the examples that fit your description of a skeptic asshole, while ignoring the fact that most of us are nice guys."

"Ad hominem, I'm afraid. You're attacking the people in skepticism, not skepticism itself. By the way, most of us are nice guys."

"Isn't that begging the question, though? Aren't you just saying 'skeptics are bad because skeptics are bad'? If you looked more closely, I think you'd find most of us are nice guys."

"You're confusing correlation and causation here. These people are not assholes because they are skeptics, they're assholes who happen to be skeptics. Most of us are actually nice guys."

"You're drawing conclusions from an insufficient sample size. If you got to know more of us, you'd realise that we were nice guys."

"Non sequitur there. You think that since these guys are assholes, all the rest of us must be. In fact, most of us are nice guys."

"Post hoc ergo propter hoc. No idea what that means, I just thought it would sound clever and impressive. You see -- most of us are actually nice guys!"

"You're being inconsistent here — it's not okay for them to insult you, but it is okay for you to insult us by association? We don't deserve that — most of us are nice guys."

"What we have here is a case of the fallacy of the excluded middle. You're only considering two extremes of skeptic — assholes or feminists — and ignoring the fact that most of us are simply nice guys."

"But absence of evidence is not evidence of absence. Just because you didn't meet any nice skeptics doesn't mean they don't exist. Most of us are actually nice guys."

Hopefully it should be clear that none of these are acceptable responses. (The fact that most of the logical fallacies are invoked incorrectly is irrelevant — and I've got to say, having compiled this list, I find it striking how easily interchangeable they all are.) This woman has confided her experiences and concerns; in these responses, you are insultingly and condescendingly attempting to diminish them, by portraying her experiences as irrelevant and her concerns as illogical. Respond like this, and you couldn't be more offensive if you said "You're too emotional, my dear: let me correct you with the firm hand of reason." In throwing around accusations of logical fallacy in such a sensitive context, you're being an asshole. And furthermore, you're excluding a voice from your privileged in-group.

The Baloney Detection Kit is a cache of offensive weapons, and for many discussions it's better to leave it behind and go in unarmed. Not every discussion is an argument, and arguments aren't always about winning. This can be a difficult concept to grasp, especially for strutting alpha male types like myself. But if we want to live in a more just and inclusive world, it's a concept that's got to sink in.