Check Them Out
- Trevor Butterworth of the Statistical Assessment Service has a short report on bogus claims of record high oil prices. Since pump prices have declined, this time the supposed record is the price of oil by the barrel. However, the inflation-adjusted price of a barrel of oil is still well below the historically highest price.
Example: "Crude Oil Price Sets Fresh Record", BBC News, 7/30/2004
Source: Trevor Butterworth, "When is a Record not a Record?", Statistical Assessment Service, 7/29/2004
Resource: Check it Out, 6/1/2004
Fallacy: Fake Precision
- The Church of Critical Thinking is on a roll, with a new "sermon" analyzing an example of one-sidedness in a CBS News report.
Source: "Issues With a Capital 'Eye'", The Church of Critical Thinking, 7/31/2004
Book Review: Influence
Title: Influence: The Psychology of Persuasion (Revised Edition)
Author: Robert B. Cialdini
Date of Publication: 1993
Review: I'm worried about this book falling into the wrong hands. Judging by the blurbs on the book jacket, what Cialdini calls "compliance professionalssales operators, fund-raisers, recruiters, advertisers, and others" (p. xii)are already on to it. For instance, The Journal of Marketing Research calls it "among the most important books written in the last 10 years" for marketers. Since the compliance pros appear to already know about it, I hope that it will start falling into the right hands.
Cialdini explains six psychological "weapons of influence" (p. xii) used to gain compliance: reciprocation, consistency, "social proof"―or, the herd mentality―liking, authority, and scarcity. These weapons can be used to get people to comply with anything from giving to a charity to torturing someone. Cialdini devotes a chapter to each of the weapons and, thankfully, ends each chapter with a section on "How to Say No". Since professionals already seem to know how to use these weapons, consumers and citizens need to learn how to defend themselves against them.
I won't discuss all of these weaponsfor that, read the bookbut I will say something about the two that have the most bearing on logical fallacies. Both of these weapons are based on shortcuts, or rules of thumb, that people use to simplify decision-making:
- "Social Proof" (Chapter 4): This is the rule of thumb of relying upon the behavior or beliefs of those aroundand similarto one as a guide to what is correct to do or believe. I don't care for the term "social proof", since this doesn't describe the process, which has nothing to do with proof. Cialdini also uses the phrase "social evidence", which is better but not very revealing. "Herd mentality" or "following the herd" would be better names for what is the psychological root of the Bandwagon Fallacy.
As Cialdini explains, relying on the behavior of those around us for evidence of what to do in a situation is often a useful rule. If you're at a dinner party, and you don't know what fork to use on the course just served, it's a good idea to look at the other diners for a clue, and harmless if you follow their example. However, as Cialdini shows through a number of historical case studies, following the herd can lead to disaster. In fact, one of the examples that Cialdini gives is that of the buffalo, which can be killed in large numbers by stampeding over a cliff. However, there is no lack of human examples:
- The infamous Kitty Genovese case: A young woman's murder was witnessed by 38 people, none of whom even called the police. No one did anything because no one else did anything.
- The Werther Effect: The suicide rate increases after a well-publicized suicide.
- The mass suicides of the People's Temple in Jonestown, Guyana: 910 people followed the leader by drinking poisoned Kool-Aid.
These examples show that the herd instinct can be frighteningly powerful. However, the situation is not hopeless, for we have a capability of reasoning that the buffalo lacks: we can learn to recognize when we're being stampeded towards a cliff.
- Authority (Chapter 6): As I explain in the entry for the fallacious appeal to authority, relying on expert opinion is not only a good rule of thumb, it is often indispensible. The world is too complicated, and there is too much to know, for us to navigate it without frequent recourse to the guidance of others. As an example of just how powerful a hold this rule has over people, Cialdini recounts the famous Stanley Milgram obedience experiments which showed that most people would rather torture someone than to disobey an authority figure. However, as with following the herd, following a leader can sometimes lead you over a cliff.
If democracy and the free market are to work as advertised, then we citizens and consumers need to learn how to defend ourselves against the compliance pros.
Check it Out
The Critical Thinking Church has a timely entry on propaganda, based on the work of the Institute for Propaganda Analysis, from the late '30s and early '40s. Some familiar fallacies put in an appearance.
Source: Church of Critical Thinking, "The 7 Habits of Highly Effective Propaganda", 7/28/2004
Non-Correlation Isn't Non-Causation
Here's law professor and "instapundit" Glenn Reynolds on why people were wrong to conclude that violence in video games causes violence in real life and that pornography leads to promiscuity:
"When teen crime and pregnancy rates were going up, people looked at things that were going onincluding increased availability of porn and violent imageryand concluded that there might be something to that correlation. It turned out that there wasn't. Porn and Duke Nukem took over the land, and yet teenagers became more responsible and less violent. Most likely, the lesson is thatonce againcorrelation isn't causation, despite policy entrepreneurs' efforts to claim otherwise.
"But regardless, the fears of the doomsayers were proven wrong. People can continue to claim that psychological research suggests that videogames lead to violence and that porn leads to promiscuity, but in the real world the evidence seems to suggest otherwise. That's an argument against regulating videogamesand it's an argument for taking other claims of impending social doom with a grain of salt."
I agree with taking doomsaying with a grain or two of salt, but Reynolds ends his article by making the very mistake that he starts out correctly criticizing. The fact that the availability of pornography and violent video games has increased while promiscuity and violence have declined is not a good argument that there is no causal relationship between them. This is because of the very complexity of such social problems that he mentions earlier: promiscuity and violence may have many causes, and though most are going down some may be rising, while the overall trend of their effects is downward. For all we know, the social ills of teenagers have been declining for different reasons, but they would have declined even more if it weren't for the bad influences of pornographic and violent media.
The most that the failure of correlation can show is that media sex and violence are not the sole causes of real-life sex and violence, but then we knew that already. If there is more than one causal factor involved in producing an effect, then no single factor may be correlated with that effect. So, "correlation isn't causation", but non-correlation isn't non-causation, either.
Source: Glenn Harlan Reynolds, "Porn and Violence: Good for America's Children?", Tech Central Station, 7/28/2004
Resource: Ice Cream Causes Rape?, 7/14/2004
I've added a new page collecting together examples of funny logical boobytraps in headlines.
Here's an interesting entry from the Church of Critical Thinking's fallacy contest:
One should keep in mind when analyzing advertisements that many ads are enthymemes with unstated conclusions. The point of an ad is to convince you to purchase the product or service advertised, so the unstated conclusion of this ad is that you should buy and drink Evian brand mineral water.
What evidence does the ad supply for this conclusion? That three-quarters of some sample of people who drank an extra liter of Evian mineral water a day had better-looking skin.
The Church's analysis of this ad correctly emphasizes the subjectivity of this evidence. However, let's put aside the question of how the subjects of the study know that their skin looks better. Instead, assume that the ad's evidence is correct, what should we conclude?
The ad gives some credence to the conclusion that if you drink Evian mineral water, your skin will look better. This may seem to be evidence that one should drink Evian, assuming that one desires better-looking skin. However, what we really need is the conclusion that your skin will look better only if you drink Evian.
In other words, the ad commits the mistake of confusing necessary with sufficient conditions. This is the confusion which underlies most of the propositional fallacies involving conditional propositions, such as Affirming the Consequent and Commuting a Conditional. The ad shows that drinking Evian is a sufficient condition for hydrating skin, but what the advertisers hope you will conclude is that it is a necessary condition, for it is only the latter which gives us a good reason to buy Evian.
Probably any brand of mineral water will hydrate your skin as well as Evian; in fact, probably nonmineral water will do so, including plain old tap water. Even fluids that are primarily water will likely work, such as fruit juices. In other words, drink an extra liter of just about any fluid and you will be better hydrated. Big deal!
In contrast, suppose that drinking Evian were a necessary condition for smoother, suppler skin. Then we would just have to go out and buy a case of Evian if we want our skin to be that way. Luckily, the ad doesn't give us any good reason to purchase an expensive mineral water, rather than just drinking more of the drink of our choice.
Source: "Contest Entry: Be Happy Naked", Church of Critical Thinking, 7/20/2004
Resource: Fallacy Contest, 7/5/2004
Michael Koplow sent in the following example:
"WASHINGTON - Rep. Chris Bell, D-Texas, who filed an ethics complaint against House Majority Leader Tom DeLay, says e-mails between Enron officials bolster his charges that DeLay illegally solicited and accepted political contributions and should be investigated. Bell said he will ask House ethics committee members to review the e-mails before deciding whether to launch a formal investigation of the Texas Republican based on the complaint Bell filed last month. DeLay has repeatedly dismissed Bellís charges, saying Bell is bitter because he lost his re-election bid in March. Republicans have said Democrats are behind the complaint. 'The last sign of a defeated and intellectually bankrupt party is a hate-filled strategy of caricature assassination,' said Jonathan Grella, a DeLay spokesman."
Mike makes the following comments on this passage:
"Two legislatorslet's call them A and Bdisagree on a piece of legislation. A accuses B of having political motivations. When you're thinking about the merits of the legislation, what should be your reaction(s) to A's accusation? I recommend two reactions:
- Who cares?, and
- I wish A would stick to the merits of the legislation and stop making ad hominem appeals.
"Imagine you agree with B's position; would you change your mind if you were able to confirm that the accusation was correct? Imagine (and this may seem far-fetched) that A and B both have political motivations; would this mean that both positions are wrong?"
In the article, DeLaythrough his spokesmangives no substantive rebuttal to Bell's charges, but instead attacks Bell's motives in bringing the charges. As Mike notes, this is an ad hominem appeal. However, no matter how bad Bell's motives might be, this has no bearing on the truth-value of his charges, which may well be true for all that. So, DeLay's response is fallacious. In our adversarial political system, we rely upon the political motivations of each side to keep the other side honest.
Also, I wonder whether Grella really said what he's quoted as saying, or the reporter just can't spell.
Source: "Enron E-mail Cited in DeLay Probe", Associated Press, 7/13/2004
- I've reformatted the front page so that the current weblog entry will appear on it.
- Update (7/17/2004): I've added a new page to the Main Menu, "Quote-Unquote", which is a compilation of quotations about fallacies and logic.
Reminder: Nobody has yet won this month's puzzle contest. Note that, despite the content of the puzzle, the winner will be the person who submits the correct answer first, not last. This is because, unlike the eccentric American billionaire who sponsored the race, I couldn't think of a way to have the final entrant win. If you're having trouble solving the puzzle, this fact is a clue!
Resource: A Patriotic Puzzle Race, 7/4/2004
Update (7/16/2004): The puzzle has been solved!
Resource: Answer to the Puzzle , 7/16/2004
Ice Cream Causes Rape?
Eugene Volokh has the scoop on a spurious correlation.
Source: Eugene Volokh, "Academic Study Reveals", The Volokh Conspiracy, 7/13/2004
Resource: Cum Hoc, Ergo Propter Hoc
Check Out What's New
I've added Julian Baggini's latest Bad Moves column as a Resource to the fallacy of begging the question.
Amazon has a new feature called a "plog", which appears to be a pseudo-weblog, consisting of personalized recommendations of books. Here's the first entry in my "plog":
"Tuesday, June 22, 2004
"My Life [Bill Clinton's autobiography] was released today; We thought you'd be interested because you bought Nonsense: A Handbook of Logical Fallacies."
The bottom of the post reads "posted by Amazon NewReleaseBot", so apparently this recommendation was generated by a programa "bot"but I don't know why it thought someone who bought a book on logical fallacies would, therefore, be interested in Clinton's book. If Amazon's bot is capable of analyzing books for logical fallacies, I'd like to know about it!
I haven't seen Michael Moore's new movie Fahrenheit 9/11, but if the ads for the movieas well as Moore's track recordare any indication, it is manipulative. Put aside the CAPITAL LETTERS and the gratuitous exclamation points! The following blurb from an ad for the movie leaves out an important qualification:
"THE YEAR'S MUST-SEE FILM! COMPELLING, FUNNY AND, MOST OF ALL, ENTERTAINING!"
Here's the quote, the whole quote, and nothing but the quote:
"The documentary's scathing attack on the war in Iraq and George W. Bush's presidency is informative, provocative, frightening, compelling, funny, manipulative and, most of all, entertaining." (Emphasis added.)
- Claudia Puig, "Put Politics Aside: 'Fahrenheit 9/11' Will Entertain", USA Today, 6/23/2004
- Ad for Fahrenheit 9/11, Indianapolis Star Weekend, 7/9/2004, p. G11
- Fallacy Watch, 4/16/2002
- Moore ad Hominems, 4/24/2002
- Moore Spinsanity, 11/19/2002
- Dissent on Moore, 6/16/2003
Brendan Nyhan has caught Joe Conason quoting a letter from President Bush to Congress about the invasion of Iraq out of context; worse still, Conason fails to include ellipses to indicate that parts of the sentence have been omitted and paraphrased. Conason quotes the letter as saying:
"I have also determined that the use of armed force against Iraq is consistent with taking necessary action against those nations who planned, authorized, committed or aided the terrorist attacks that occurred on September 11."
The actual sentence in the letter is as follows:
"I have also determined that the use of armed force against Iraq is consistent with the United States and other countries continuing to take the necessary actions against international terrorists and terrorist organizations, including those nations, organiza-tions [sic], or persons who planned, authorized, committed, or aided the terrorist attacks that occurred on September 11, 2001." (Emphasis added.)
Omitting the italicized parts, especially the references to organizations and persons, makes it seem that Bush is holding some nationspresumably including Iraqresponsible for the 9/11 attacks. In contrast, the complete sentence is simply covering all bases.
A second logical point has to do with the word "consistent", which seems to have confused many commentators on this issue. To say that two propositions are consistent is simply to say that both could be true. Also, it's important to realize that Bush's letter was a Presidential determination as required by the Congressional joint resolution authorizing the use of military force against Iraq:
"In connection with the exercise of the authority granted to use force the President shall, prior to such exercise or as soon there after as may be feasible, but no later than 48 hours after exercising such authority, make available to the Speaker of the House of Representatives and the President pro tempore of the Senate his determination that acting pursuant to this resolution is consistent with the United States and other countries continuing to take the necessary actions against international terrorists and terrorist organizations, including those nations, organizations or persons who planned, authorized, committed or aided the terrorists attacks that occurred on September 11, 2001."
This part of the resolution was apparently added by those members of Congress who were worried that a military invasion of Iraq would hamper anti-terrorism efforts in Afghanistan and elsewhere. Congress, therefore, required that in order for military force to be authorized, the President had to determine that it was "consistent" with such anti-terrorism actions, that is, that the U.S. could both fight a war with Iraq and fight terrorism, too. In other words, neither the congressional resolution nor Bush's determination are claiming any link of Iraq to 9/11.
- George W. Bush, "Presidential Letter", The White House, 3/21/2003
- Joe Conason, "9/11 Panel Becomes Cheneyís Nightmare", New York Observer, 6/28/2004
- "Joint Resolution to Authorize the Use of United States Armed Forces Against Iraq ", The White House, 10/2/2002
Via: Brendan Nyhan, "Conason Botches Bush's Letter to Congress", Spinsanity, 7/5/2004
Update (7/8/2004): In today's column, Conason corrects this contextomy.
Source: Joe Conason, "Nader Returns, With G.O.P. Help", New York Observer, 7/8/2004
I didn't realize that it's an organized religion, but the Church of Critical Thinking is holding a contest this month with a prizeor, more accurately, a boobyprizefor the person who finds the "best" fallacious argument in the media. I will sit out this contest as a semi-professional fallacy hunter, andmore importantlyI really don't want that prize!
Source: "Announcing the First Church of Critical Thinking CONTEST!", The Church of Critical Thinking, 7/1/2004
A Patriotic Puzzle Race
An eccentric American, who had made billions in the computer industry, wanted to challenge the predominance of the Tour de France in bicycle racing. So, he offered a prize of one million dollars to the winner of his race, which was to take place in the United States. The eccentric thing about the race was that the winner was to be the cyclist whose bike came in last. Everyone was puzzled by how such a race could possibly work: all of the cyclists would just wait around at the starting line for the others to go. The race would never get started, so it could never end. How did the clever American solve this problem?
The first person to send the correct answer to the Fallacy Files wins the puzzle race!
In addition to its regular polling, Newsweek magazine occasionally does what it calls a "GENEXT" poll of young voters. These polls usually use a smaller sample than others; for instance, other Newsweek polls typically have a sample of over a thousand, whereas the GENEXT polls are about a third of that size. This may well be because the questions targeted at younger voters are asked during the full poll, so that the GENEXT sample is a sub-sample of that of the larger poll.
In any case, the sample size of a poll and its margin of error vary inversely, that is, the smaller the sample, the larger the error. So, whereas a full Newsweek poll usually has a margin of error of ±3%, the GENEXT polls' margin of error is ±5.1%. For this reason, the GENEXT polls are usually not sensitive enough for their results to be meaningfulthough they are certainly more meaningful than the completely worthless online polls that always accompany the reports of the poll results. With that background in mind, here is how Newsweek reports the main results of today's GENEXT poll:
"The most noteworthy shifts in young voter opinion are on international issues. In May, more than half (55 percent) of voters 18 to 29 supported the presidentís handling of foreign policy issues and the war on terrorism, but since then his numbers have reversed. Now just 47 percent approve and half of the respondents say they (52 percent) disapprove. More potentially challenging to the Bush campaign is the fact that more than half of young voters (60 percent, up from 55 percent in May) now disapprove of the presidentís handling of Iraq ."
Given the large margin of error on both of these polls, all of these results are actually statistically insignificant, and represent no real shift in opinion among young voters. For example, the "change" in support for the President's foreign policy supposedly dropped from 55% in May to 47% currently. However, given the margin of error, the results could be off by as much as 5%rounding the margin of error to the nearest whole percentage point. So, the May result could have been as low as 50% and the present one as high as 52%, which is actually an increase!
Source: Brian Braiker, "Big Man on Campus", Newsweek, 7/2/2004
Resource: Poll Watch, 3/21/2004
Fallacy: Fake Precision
Source: "Headlines", The Tonight Show
Answer to the Puzzle (7/16/2004): And we have a winner! John Congdon was the first to send in the correct answer. WARNING: Solution ahead! Here is the correct solution in John's words:
"He [the eccentric American billionaire] had all the racers switch bikes. Since the winner is the one whose bike comes in last (not the rider who comes in last), all the riders would try to be across the finish line before their own bike."
Notice that the statement of the puzzle does not say that the racers have to ride their own bikes, though this is the usual thing to do, and is so common that one is not likely to think of this alternative. Honorable mentions go to two runners-up. Lawrence Mayes offered the following solution:
"The billionaire should offer one million and one dollars to the cyclist who comes in second from last, one million and two dollars to the one who comes in third from last and so. To keep solvent he would need to restrict the number of entrantshe could do so by limiting each country to one competitor. If there were no more than a hundred competitors, say, there would be less than a hundred dollars difference between coming in first and 'winning'. So there would be a huge incentive to finish but almost none to tarry."
Lawrence cleverly takes advantage of the fact that the statement of the puzzle does not rule out monetary prizes for the other contestants. However, this solution has the disadvantage that it seems strange to call the rider who gets the smallest prize the "winner". Jim McSharar sent in another clever answer:
"The race ends when the first bike touches the finish line. Since there is only one prize, nothing behind first counts so the race is over and the winner is also last."
Nothing in the puzzle rules out the last bike also being the first one, because it is the only one across the finish line. I'm unsure whether it's an advantage or a disadvantage, but it's hard to see how this race would differ from an ordinary race!
Congratulations to the contestants!
Source: Martin Gardner, aha! Insight (W.H. Freeman, 1978), p. 107.