As a long-time and frequent participant in various kinds of online debate, I've encountered many instances of certain logical fallacies. As a direct consequence I've become quite good at identifying them, not only as they appear in textbooks but also as they appear in the much messier internet. Sometimes, when I identify a fallacy by name, the people who have committed the fallacies - who have been quite deliberately using them to advance their own position - act as though it doesn't matter, that logic is just for pointy-headed academics and of no use in a real conversation. That belief is wrong. When people disagree, logic is as essential to understanding and possibly resolving that disagreement as language itself. A fallacy is a form of untruth whether it occurs in a presidential debate or a hallway chat, and there's nothing pointy-headed about identifying untruth as such. This article is my attempt to bring the identification of certain common fallacies down to earth.
Argumentum ad hominem is one of the two fallacies that does seem to be widely known on the net (the other is strawman). Literally, it means "to the man" and it refers to discrediting the source of information rather than the information itself. One of the important rules of logic is that a fact doesn't cease to be a fact because of who stated it or why, and an ad hominem argument violates that rule. "Poisoning the well" is a preemptive form of ad hominem attack in which someone tries to impugn the source of contrary information before a claim is even made, as in "of course, a communist would tell you something different" and similar statements.
The one case in which it's appropriate to focus on the source of a statement rather than the statement itself is when the source itself is treated as proof of the statement. This type of argument is called an appeal to authority (or argumentum ad verecundiam) and is sort of the mirror image of argumentum ad hominem - it assumes that something is true instead of false because of who said it. At least half the time when you try to refute an appeal to authority, you will be accused of an ad hominem attack, unless you very carefully retain focus on the statements themselves, and make it clear that the identity of the source makes something neither true nor false.
Everybody knows the strawman. Everybody, it seems, loves the strawman. A strawman is a statement or belief attributed to someone else for the purpose of refuting it, but there's actually more to it than that. If a strawman is obviously unrelated to someone's actual argument, it's not effective at all. To be effective, a strawman must resemble the argument one is (fallaciously) trying to refute. In other words, it's an attempt to disprove a real conclusion by disproving a fake one that looks similar. This is very closely related to the lesser-known (but just as often practiced) tactic of disproof by fallacy, which seeks to invalidate a real conclusion by refuting one particular argument leading to that conclusion. If I argue that 2+2=4 because the moon is made of green cheese, my argument is easily refuted but the conclusion is still true. You could refute a hundred such arguments and the conclusion would still be true. The only way to refute the conclusion itself is to present an alternative conclusion, not to attack the arguments behind the first.
This is another favorite on the net: just keep saying something over and over again until all of your opponents give up. Ignore any contrary evidence, or identification of internal contradiction, as though they had never been posted. Being offensive as well as stubborn will result in even greater attrition and bring victory even sooner. Sound familiar? Practitioners of this technique are legion, and the best response is to ignore them. Let them have the last word; have faith that others can tell the difference between obstinacy and truth.
There are several closely related fallacies based on the often-made and often-mistaken assumption that anything which is not A must be B. This is true in some cases - e.g. heads and tails - but the possibility of other alternatives should always be considered. If A is the left side of the road (or political spectrum) and B is the right, the "excluded middle" in one variant of this fallacy is the middle of the road, or perhaps the line one-third of the way from the left. Bifurcation can occur even when there's no linear relationship between options, though; if the fact that I'm not wearing blue underwear is used to "prove" that I'm wearing red underwear instead, perhaps the excluded middle is plaid.
In a similar vein, the "slippery slope" assumes that as soon as we move away from A we will inevitably arrive at B - if not right away, then eventually. If we allow a sales-tax increase to 5% it's only a matter of time before it's increased again to 10%, the argument goes. The "camel's nose" is another metaphor for the same thing, based on the idea that once a camel gets his nose into your tent it's only a matter of time before the rest of him will follow. As with the excluded middle, the possibility of options in between the extremes is ignored. Slopes are never perfectly slippery, as any child who has gone sledding could tell you, and sometimes the camel's nose actually is the only thing that gets into the tent.
One tricky aspect of bifurcation is that you will often be accused of it when you attempt to refute other kinds of flawed arguments. For example, just about any attempt at reductio ad absurdum, which attempts to refute an argument by following it to its logical conclusion, can bring howls of "excluded middle" from the target. The key in such cases is to distinguish between a logical connection ("if A then anything but B would be inconsistent") and a causal one ("if A happens then B will happen too").
If you can't win an argument, make it impossible for your opponent to win either. That's the basic philosophy behind these tactics, which are almost always used together. Instead of trying to prove X myself, I embark on a three step process:
There are problems in all three stages of this procedure. The first is obviously an example of bifurcation as discussed earlier. The third (the "moving the goalposts" part) is simply a way to waste my opponent's time, and indeed this technique is often used more to sink a debate into oblivion than to win it outright. The second stage is the most interesting one, because it's natural to ask where the burden of proof really should lie. There's no absolute rule, unfortunately. One common rule is to say that the burden lies with whoever made the first claim, but that can lead to a form of gamesmanship in which each person tries to insinuate their beliefs but doesn't make any outright claims, and the first person who does make a claim gets jumped on. My personal favorite approach is to apply Occam's Razor, to accept the argument that requires the fewest entities or premises by default and place the burden of proof on any alternative.
Latin for "you too"; this is the fallacy of assuming two wrongs make a right. If you're arguing that George W. Bush is a danger to the world, any argument of the form "yeah, but Bill Clinton was even worse" qualifies. There shouldn't be any need to say any more, but this fallacy (often in exactly the form of the example) occurs amazingly often.
This last family of fallacies all revolve around the original Latin phrase "post hoc ergo propter hoc" meaning "after this therefore because of this". If most college professors are liberal, which of being liberal and being a professor is the cause and which is the effect? There are not two, but three distinct possibilities:
In short, never take anyone's assumption of a cause/effect relationship at face value. People often try to slip such assumptions into a debate without calling attention to them, and it's much easier to address those assumptions when they're first made than later.