Q: I am a medical student with a very random and growing interest in fallacies and paradoxes. I have spent much time looking for recommendations on your website (among others) for a book (textbook or otherwise) that covers most, if not all, formal and informal fallacies. Do you know of any such books that would cover the fallacies in a manner similar to what you have done on the taxonomy portion of your website? Or are there any other books you would recommend?―Michelle Taylor
A: There is no book that uses a taxonomy, as far as I know―and I know pretty far. Of course, every logician since Aristotle who has written anything much on the subject has classified fallacies in some fashion, but these systems of classification are comparatively "flat", and lack the tree-like structure of the taxonomy. Instead, fallacies are usually classified in a few broad categories, such as "fallacies of relevance", "linguistic fallacies", "fallacies of presumption", etc.
Most books that cover a large number of fallacies tend to restrict themselves to informal fallacies, and formal fallacies are often only dealt with in passing in books on formal logic. Many introductory logic textbooks will cover both types of fallacy, but usually only about a dozen to two dozen informal fallacies. Moreover, the informal fallacies covered in textbooks are the "usual suspects", that is, the traditional logical fallacies many of which go back to Aristotle: begging the question, ad hominem, straw man, etc. You probably won't find any of the fallacies discovered in recent decades by psychologists―such as the base rate fallacy, the conjunction fallacy, and the hot hand fallacy―or those studied by statisticians―such as the regression fallacy, the Texas sharpshooter fallacy, and the multiple comparisons fallacy.
That said, there are many fine textbooks from which you can learn not only much about fallacies, but also the positive principles of logic. Formal fallacies, in particular, can only be fully understood within the context of the system of formal logic in which they occur; for instance, to fully appreciate syllogistic fallacies requires an understanding of the formal system of categorical syllogisms.
The closest thing to a single book that meets your criteria is Madsen Pirie's The Book of the Fallacy, which is out of print, hard to find, and expensive when found. Pirie's more recent book How to Win Every Argument: The Use and Abuse of Logic appears to be an expanded version of his earlier work, so I think that it is safe to recommend even though I haven't read it yet.
The Puzzle of the Terrorist Acquaintance
The Agency for Counter-Terrorism (ACT) comes to you with a logical problem. The agency has information on four subjects, one of whom is known to be a terrorist. To protect the innocent, we will refer to them as Subjects 1 through 4.
Subject 1 is the known terrorist. Subject 1 is acquainted with Subject 2. Subject 2 has met Subject 3, therefore they are acquainted. Subject 3 was observed talking to Subject 4, so they're also acquainted. Subject 4 is known not to be a terrorist. It's not known whether Subjects 2 and 3 are terrorists or not.
ACT wants to know: among the four subjects, is a terrorist acquainted with a non-terrorist? Choose an answer:
- There is not enough information to determine whether a terrorist and a non-terrorist are acquainted.
Acknowledgment: Thanks to Vasilios Magriplis for pointing out a typographical error in the original wording of the puzzle that has now been corrected.
Are you intelligent but irrational?
Test yourself with an article in the latest issue of Scientific American Mind on the difference between intelligence and rationality by psychologist Keith Stanovich. It contains a number of puzzles that may make you feel foolish, but don't feel too bad if you get a wrong answer: you're in good company. Regular readers of The Fallacy Files should recognize the Wason selection task and a puzzle based on the base rate fallacy, and I hope would not be fooled by them.
I've known a lot of highly intelligent people, including some much smarter than I am. However, some of them were perplexingly irrational, at least about some things. I used to be very puzzled by this, until I drew a distinction between intelligence/stupidity on one hand, and wisdom/foolishness on the other. It's perfectly possible to be an intelligent fool: I've known a few! It may even be possible to be a wise idiot, though I've never actually met one. However, the point is that raw intelligence and rationality don't always go hand-in-hand. Intelligence is innate, but rationality must be cultivated.
The article is based on Stanovich's book What Intelligence Tests Miss: The Psychology of Rational Thought, which I would love to receive a review copy of.
Source: Keith E. Stanovich, "Rational and Irrational Thought: The Thinking That IQ Tests Miss", Scientific American Mind, 11/2009
New Books: Denialism and Unscientific America
Continuing the twin themes of "where's the harm?" and weird science, here's a couple of new books: Chris Mooney and Sheril Kirshenbaum's Unscientific America and Michael Specter's Denialism. Both appear to deal with the widespread scientific illiteracy that helps lead to the harm that I've noted in previous entries. For instance, the quantum quackery practiced by James Arthur Ray is made possible by the fact that so many people have no idea what quantum mechanics is all about, and can't tell the difference between the real thing and bafflegab. You don't have to be a physicist to be able to tell that Rhonda Byrne doesn't know what she's talking about; you just need to be scientifically literate. I haven't done a book club in a long time, and it's possible that one or both of these books might make good material. As always, it would be nice if someone would send me review copies.
- Chris Mooney & Sheril Kirshenbaum, Unscientific America: How Scientific Illiteracy Threatens our Future (2009)
- Michael Specter, Denialism: How Irrational Thinking Hinders Scientific Progress, Harms the Planet, and Threatens Our Lives (2009)
Where's the Harm?
- (11/7/2009) Also here:
Despite major bombings that have rattled the nation, … Iraqís security forces have been relying on a device to detect bombs and weapons that … technical experts say is useless. … The small hand-held wand, with a telescopic antenna on a swivel, is being used at hundreds of checkpoints in Iraq. But…a retired United States Air Force officer, Lt. Col. Hal Bidlack,…described the wand as nothing more than an explosives divining rod. … The suicide bombers who managed to get two tons of explosives into downtown Baghdad on Oct. 25, killing 155 people and destroying three ministries, had to pass at least one checkpoint where [it] is typically deployed,….
The Iraqis have spent 85 million dollars on these pieces of junk. If you want a dowsing rod, you can make one yourself from the branch of a willow tree, or from a wire coat hanger. However, you won't be able to detect explosives with it any better than by flipping a coin, and people will be blown up as a result, but at least you won't be wasting millions of dollars. Apparently, for less than 5 million dollars they could equip every checkpoint in Baghdad with a bomb-sniffing dog.
Dale Murray, head of the the National Explosive Engineering Sciences Security Center at Sandia Labs, which does testing for the Department of Defense, said the center had "tested several devices in this category, and none have ever performed better than random chance." … The Justice Department has warned against buying a variety of products that claim to detect explosives at a distance with a portable device. … "I donít care about Sandia or the Department of Justice or any of them," General Jabiri said. "I know more about this issue than the Americans do. In fact, I know more about bombs than anyone in the world." He attributed the decrease in bombings in Baghdad since 2007 to the use of the wands at checkpoints.
It's time to play name that fallacy! For the answer, see below.
If, as often happens, no explosives or weapons are found, the police may blame a false positive on other things found in the car, like perfume, air fresheners or gold fillings in the driverís teeth. … During an interview on Tuesday, General Jabiri challenged a Times reporter to test [it], placing a grenade and a machine pistol in plain view in his office. Despite two attempts, the wand did not detect the weapons when used by the reporter but did so each time it was used by a policeman. "You need more training," the general said.
This kind of excuse-making by ad hoc hypothesis is a standard feature of pseudoscience: a hit is counted as a hit, but a miss is not counted as a miss―instead, it's excused and explained away. Every "psychic" detective does the same thing by throwing out lots of information about where a missing person will be found, then counting only the hits and ignoring the misses. With this kind of scoring system, you can't lose!
That's why it's necessary to test dowsing and "psychics" in such a way that all the hits and misses are counted. When that is done, dowsers do no better than chance. Of course, the dowsers always have explanations for their failures: conditions weren't right, something was interfering with their powers, the tests were rigged against them, etc. Similarly, the General―whose idea it was for an untrained reporter to test the device!―excuses its failure as due to the reporter's lack of training.
The Iraqis could do just as well by using some kind of randomizer at checkpoints to pick out cars for searches as they can with these high-priced dowsing rods.
Source: Rod Nordland, "Iraq Swears by Bomb Detector U.S. Sees as Useless", The New York Times, 11/3/2009
Fallacy: Post Hoc
Acknowledgment: Thanks to Vance Ricks for calling this to my attention.
- Here it is:
The shortage of swine-flu vaccine results not from drug-company greed or outsize demand but almost entirely from the government's decision to pander to unfounded and unscientific fear. As The Wall Street Journal reported last week, the US government set out to have the H1N1 vaccine produced largely in single-dose syringes―a demand that has set back production considerably, because multidose vials are far easier to make. And the only reason to seek single-dose production was to please people needlessly worried about the preservative thimerasol, which is used to provide multiple doses of the vaccine. The fear―utterly groundless and repeatedly debunked is that thimerasol can cause autism and other neurological disorders in infants and other young children. If not for that decision, we'd have more than enough vaccine. Instead, because the government yielded to pressure from antivaccine fringe groups, we're behind the curve on protecting millions of children from swine flu.
So, some children will get the flu and some may even die from it because of fallacious fears that vaccines caused other children's autism. It's hard to say how much blame for this should fall on celebrities such as Jenny McCarthy and Bill Maher, but it's definitely nonzero. Of course, as much or more blame must fall on those who listened to these celebrities and took their silly advice seriously.
- Robert Goldberg, "Why you can't get swine-flu vaccine", The New York Post, 11/3/2009
- Tara Parker-Pope, "Bill Maher vs. the Flu Vaccine", The New York Times, 10/13/2009
The Multiple Comparisons Fallacy. I haven't added it to the Taxonomy yet; that is to come. As usual, if you notice any errors or omissions, please let me know.
Solution to the Puzzle of the Terrorist Acquaintance: Yes, a terrorist is acquainted with a non-terrorist. To see this, consider Subject 3. Either Subject 3 is a terrorist or not. Suppose that Subject 3 is a terrorist; then a terrorist is acquainted with a non-terrorist, since Subject 3 is acquainted with Subject 4. What if Subject 3 is not a terrorist? Then consider Subject 2, who is also either a terrorist or not. If Subject 2 is a terrorist, then a terrorist is acquainted with a non-terrorist, since Subject 2 is acquainted with Subject 3. So, suppose that Subject 2 is not a terrorist; then a terrorist is acquainted with a non-terrorist, since Subject 1 is acquainted with Subject 2.
Another way to solve the puzzle would be to consider all the possibilities for Subjects 2 and 3: they're both terrorists, neither is a terrorist, Subject 2 is a terrorist but Subject 3 is not, and Subject 2 is not a terrorist but Subject 3 is. In each of these possibilities, a terrorist is acquainted with a non-terrorist.
If you read the article discussed in the previous entry, you may have noticed that this puzzle is a variant of the first puzzle given in the article. I wondered what would happen if there were four people rather than three; specifically, I wondered whether it would still be solvable. As you can see, it is. Of course, I also changed the relation between the characters from looking to being acquainted, and the property of being married to being a terrorist.