Information icon.svg Results for the 2024 RationalWiki Moderator Election have now been posted. Thank you for participating in this election, and congratulations to the winners!

Essay:Argumentum abusi fallacia

From RationalWiki
Jump to navigation Jump to search
Essay.svg This essay is an original work by Armondikov.
It does not necessarily reflect the views expressed in RationalWiki's Mission Statement, but we welcome discussion of a broad range of ideas.
Unless otherwise stated, this is original content, released under CC-BY-SA 3.0 or any later version. See RationalWiki:Copyrights.
Feel free to make comments on the talk page, which will probably be far more interesting, and might reflect a broader range of RationalWiki editors' thoughts.

Argumentum abusi fallacia (Latin for "argument of the abused fallacy") is the incorrect use of a formal or informal logical fallacy. As there is an absolute myriad of fallacies to choose from, it's quite easy to not be entirely familiar with them all. False accusations that someone is making a fallacious argument, when they're not, become common. Someone might be making a bad or incorrect argument, but there are ways to be wrong without being fallacious about it and there are ways to be right while still being fallacious. There are also ways to counter an argument without providing a list of fallacies - in the business we call these things "counter" arguments. The use of argumentum abusi fallacia then becomes something of an argument by assertion, in which an attempt is made to refute an argument simply by citing the name of a fallacy, without any further explanation of why.

This is often a result of falling into skeptic jargon, or just discovering that there are terms not just to describe that someone is wrong but why they're wrong. So, thanks to convenience and in an effort to show off this new-found knowledge - and, hey, some of it's in Latin! - someone might be very keen to drop the name of a formal or informal fallacy into conversation. However, just because you can reduce why someone is wrong to a single fany name, doesn't mean the names of fallacies can be thrown around at will and automatically be correct. Arguments need to be demonstrated as being fallacious, and usually this is possible without using any Latin at all. Unless you're attempting to answer a Gish gallop in real time, in which case you might not have much choice but to just go for the shorthand. So long as you use the shorthand correctly, of course.

Formal and informal[edit]

In short, a formal logical fallacy is a fallacy in the structure of an argument. If you distil the argument down into symbolic logic, which is the quasi-mathematical way of representing arguments and implications, a formal fallacy is one where the pieces simply don't fit together. Because this is all represented symbolically, what those letters actually mean don't count. In a formal statement like "PQ, P, therefore Q" it doesn't matter what P and Q stand for. The formal fallacies are rarely misused; though mostly because it's far easier to take issues with content and style than to break an argument down to a logical form and track down the errors - and if one was to do that, then the fallacies become evident and difficult to misinterpret or produce a false positive. For instance, accusations of a non sequitur argument are usually valid, since that disconnect is easy to spot and demonstrate.

Informal logical fallacies are issues with the content of an argument. I.e., what the P and Q actually stand for. Unlike the formal logical fallacies, the informal fallacies are rife with pitfalls where they are misused, creating a ton of distracting false positives. At worst, their misuse comes from an inability for someone to create a new argument of their own or to address an opponent's points; they simply accuse them of making a fallacy and have done with it.

The examples below chart the most common offenders.

Example: Special pleading[edit]

The cosmological argument as originally formulated by St. Thomas Aquinas read roughly as follows:

Everything that exists requires a cause. There cannot be any infinite regression of causes. Therefore, the universe must have a first cause: God.

The trouble with this argument is that it exhibits special pleading. God is arbitrarily, without any supporting reason, exempted from the requirement that "all things require a cause". Often the response to this argument is "why can't the universe be acausal if God can be?" and thus the special pleading begins. Either God is made an exemption "by definition" (a very strong case of special pleading) or other wild assertions are made that make God "not count". This is a legitimate fallacy: God needs a real reason to be exempted from the premises of the arguments.

So, in response to this, the argument has been modified slightly. This is the Kalām cosmological argument, touted ad nauseam by William Lane Craig, which usually denotes this slight alteration over the Thomistic version above:

Everything that begins to exist requires a cause. There cannot be any infinite regression of causes. Therefore, the universe must have a first cause: God.

If somebody were to dismiss this modified version of the argument with "that's special pleading!" it would be an incorrect use of the fallacy. This is is because the modification explicitly exempts an "eternal" God from the need for a cause, so does not contain any special pleading. The underlying logic is on better and less fallacious grounds as "special pleading" is a formal logical fallacy - a problem with the structure, not content. Special pleading is stating , and giving ad hoc exceptions. If these exceptions are, in fact, built into the logical conditions then there is nothing "special" about the pleading, ergo, no fallacy.

Example: Ad hominem[edit]

An ad hominem argument, from the Latin for "to the man", is something that doesn't attack the questions and points at hand, but the messenger and the person delivering it. This is one of the most common misuses of an assertion of fallacy because people can all too easily confuse ad hominem with "crass insult". If someone calls someone else a total douchebag prick out of the blue in a discussion, it might be mean, it might be unproductive, hell it might even be correct, but it is not necessarily fallacious. Ad hominem attacks dismiss an argument because of a completely unrelated property of the person delivering the argument.

Consider the following:

The war in Iraq is illegal because George W. Bush is an incompetent buffoon.

This sort of thing is an ad hominem attack because, buffoon or not, Bush's intellect has little bearing on the validity of a war. However, we can rephrase this slightly.

George W. Bush is an incompetent buffoon because the war in Iraq is illegal.

This is on slightly firmer ground. In short, while it might still be irrelevant, Bush's lack of brain isn't being used to justify a legal status, and it's easily arguable and conceivable that someone engaging their country in an illegal war could be described by the words "incompetent" and "buffoon" amongst several others. Yet, if faced with no alternative rebuttal, this might well be accused of being an ad hominem attack, just because it makes an egregious personal insult.

(Note: the two above are forms the same ideas with the implication reversed: and . For a treatment of why these aren't the same statement, see affirming the consequent and its statistical cousin confusion of the inverse.)

A few others[edit]

Here is a brief run-down of a few other pieces of skeptical jargon that occasionally get misused:

Argument from adverse consequences
This is the fallacy that states that because X is unfavourable, or would cause problems, X isn't true. It is fallacious because the effects of a hypothesis have no bearing on its truth value - objective reality cares little about whether it screws us over. However, there are cases where there is no "truth value" to be found. For instance, social policy. Here, consequences of a decision are all we have to determine a correct course of action - in terms of setting a policy or making a decision, the consequences are a stand in for the "evidence" you would expect to see from a testable hypothesis.
Appeal to authority
As noted in the article, appeal to authority is perfectly valid when it's an appeal to a relevant and experienced authority. Talking to Brian Cox or Stephen Hawking about physics, and Richard Dawkins about evolutionary biology, for example is fine because they have studied and contributed to these subjects. While it is true that even in these cases their arguments need to be judged on their own merits and aren't correct because they are authorities, this fallacy is often mistaken as a carte blanche to ignore anything said by an authority.
Begging the question
People often use this to mean "raises the question"... but that's something else entirely. Circular logic can often be easily confused with just normal logic as any individual logical step should be so undeniably sound that it might seem, to the untrained eye, just to state the obvious. Consider PQ, P, therefore Q, for example - this is formal logic working at its finest, but "P, therefore Q" might very easily appear circular.
Confirmation bias
Remember, just because you've found evidence that still supports your hypothesis, doesn't mean you are necessarily guilty of confirmation bias. The standard model of particle physics has withstood a lot of testing, but those testing it aren't guilty of confirmation bias; which is a description of how you go about searching for that evidence by building experiments and tests that can only prove your point and may never fail. Specifically, whether you bother to look for something that will disprove your point. The Wason card problem illustrates the point of a confirmation bias nicely: if you first seek out evidence to disconfirm the hypothesis of the problem, you can solve the problem in fewer card turns and the falsifying test always needs to be performed, while the confirming test is irrelevant to the hypothesis given.
Correlation does not equal causation
Often, this can be cited in a way that almost completely denies the correlation too. If two events (P and Q) correlate significantly, then the probability of you having P when you have Q is still higher than when you don't have it. A lack of causality between still won't stop this if the correlation is experimentally verified. For instance, in the textbook example of shoe size correlating with reading ability, the main cause is that both correlate quite nicely with age - but unless you control for age, experience and education, the correlation between reading ability and shoe size will still exist. Indeed, more precisely determining causality from data is done this way; by controlling for confounding variables to see what relationships remain.
Equivocation
Equivocation is effectively an illusion of language only - the fact that words can have multiple meanings and definitions in different contexts and so might get confused. It's a simple mistake, really. For example, conflating biological evolution (Darwinian natural selection) with stellar evolution that tracks the lives of stars. Sometimes people will call "equivocation" when someone is using an analogy - but providing the person making the argument by analogy knows what they're saying and where the analogy works and where it fails, there is certainly no equivocation involved.
Extraordinary claims require extraordinary evidence
This quip has often been used to describe how believing in ghosts requires a substantial degree of evidence because such a concept would go against what we already know - it would rattle the whole of science far more than merely not finding the Higgs boson would. But sometimes it can produce a straw man that's used to dismiss any and all evidence for something because it simply isn't miraculous enough. Homeopathy would just need a statistically significant improvement over a meta-study to prove it works for a particular illness, it wouldn't need to magically cure every cancer it's tried on to prove itself (the fact it does neither is beside the point, it's about correctly judging what evidence is required).
Fallacy fallacy
Check this comic for a moment - the basic idea is there, that merely listing a fallacy doesn't necessarily disprove something (you can be right while being fallacious, or you can be wrong without). However, the fallacy fallacy doesn't really apply if you can find several, and more from a different approach, and the objections also involve fallacies. In that case, the odds are the argument is, indeed, utter bullshit. Claiming the "fallacy fallacy" here could be the... fallacy fallacy fallacy? The slightly newer adage about correlation and causation is relevant here; correlation [between your conclusion and fallacious arguments] does not imply causation, but it does wiggle its eyebrows suggestively saying "look over there".
Slippery slope
"If you allow X, then Y and Z are certain to follow - therefore X is bad." This is only an invalid argument under certain conditions. Specifically, how realistic is the slope involved here? Raising house prices on one property is likely to cause a race to increase the prices of the surrounding houses, this slope is fairly realistic. Gay marriage leading to the legalisation of bestiality is significantly less realistic on account of the principle of informed consent. This is tied up with argument from adverse consequences.
Straw man
A straw man argument is one where a person deliberately (and perhaps knowingly) sets up a false version of what they're arguing against in order to defeat it. Straw men are often seen in protracted arguments where a particular "party line" exists; for instance, the crocoduck as a supposed "example" of evolution - straw man versions of creationist beliefs appear equally often, particular amongst parodists. However, true straw man arguments tend to exist mostly when speakers or writers have had time to process what the other has said, and the straw man is really identified as such when they continue to use those arguments despite repeated corrections. The misuse of the straw man idea appears in a few different ways.
  1. What tends to happen in quick-fire debates is that someone will accuse the other of forming a straw man argument, when really all that has occurred is a misunderstanding - and 90% of the time, the responsibility for the misunderstanding is the person doing the talking, not the listening. Accusing an opponent of forming a straw man argument forms a tactic to deflect people from noticing that the point may not have been laid out very well to start with.
  2. A "straw disclaimer" can exist in which someone is saying that they're arguing for x but specifically say they're not arguing for a slight modification of x. If a counter-argument suggests that there is no practical difference between the two, it's very easy to avoid this argument completely by shouting "straw man". This often crops up when defending mentioning slavery in the Bible; as the Biblical rules refer to how to treat fellow Hebrews and don't refer to "chattel slavery" - even though, for most practical and moral purposes, there's not a tremendous amount of difference.
  3. As a response to using a reductio ad absurdum where someone is shown to endorse a position they didn't explicitly endorse initially. While a reductio ad absurdum argument can produce a straw man, this isn't always the case and the correct response is to demonstrate the fallaciousness of the reductio argument, not to dismiss its conclusions outright.

See also[edit]