Previous Month | RSS/XML | Current | Next Month

September 26th, 2006 (Permalink)

New Book

The latest issue of the Skeptical Inquirer magazine has a favorable review of a new book called "Don't Believe Everything You Think: The 6 Basic Mistakes We Make in Thinking", by Thomas Kida. The review isn't yet available on the web―and may never be, since SI doesn't make all of its articles accessible from its website―though it can be downloaded in digital form from Amazon. According to the review, the book discusses six cognitive mistakes, including the anecdotal fallacy and confirmation bias. The review mentions that hindsight bias and the bandwagon fallacy also come in for some attention. I may review the book in full in the future, especially if the publisher will be so kind as to send me a review copy.

Source: Peter Lamal, "Why What We Know isn't Necessarily So", Skeptical Inquirer, September/October, 2006, p. 55.

Resource: Thomas Kida, Don't Believe Everything You Think: The 6 Basic Mistakes We Make in Thinking (2006)

September 24th, 2006 (Permalink)

Ad Festivus for the Rest of Us

Mark Ellis of the Mathematics Division of Central Piedmont Community College sends the following fallacy suggestion:

I have informally (no pun intended) thrown about a new name for a fallacy. I've quickly glanced at your list and others and have yet to find one that meets this one: My submission for your approval is "ad festivus" meaning "to the clown." Now, my Latin is deplorable, but I think I remember reading in one of Copleston's volumes a footnote mentioning that "festivus" meant or was a synonym for "clown" or "fool." My idea for ad festivus would then be: Appealing to the conclusions of another simply because they make you laugh or feel good. A subfallacy of the ad verecundiam, I'm sure.

While it's uncommon to list a fallacy involving humor, it's not unknown. T. Edward Damer lists a fallacy of "Resort to Humor or Ridicule", and Robert J. Gula briefly mentions the use of humor, sarcasm, ridicule, and wit. Both treat the fallacious use of humor as a diversionary tactic, rather than an appeal to authority, since laughter can easily cause the audience to lose track of the argument.

According to my Latin dictionary, "festivus" just means "festive" or "merry". The closest word that I can find for "clown" in Latin is "rusticus", from which the English word "rustic" is derived. Apparently, sophisticated Romans found country folk humorous, as do some urbanites today. So, an alternative name for the fallacy would be "argumentum ad rusticum", or something along those lines. Perhaps a Latin literate reader will send in a better suggestion. However, it isn't necessary to give every fallacy a Latin name, though "ad rusticum" is preferable to Damer's bland and pedantic "Resort to Humor or Ridicule".

I have some doubts, however, that the appeal to humor should be considered a logical fallacy. Jokes are usually not arguments at all, and therefore not fallacious arguments, a fortiori―to throw in a little more Latin. No doubt humor and ridicule can be used as rhetorical dirty tricks to distract the audience, as well as to put the ridiculed side at a disadvantage in a debate. I assume that your suggestion that it would be a subfallacy of appeal to misleading authority is based on the psychological fact that we are likely to be swayed by people who make us laugh. People may be more likely to trust a Rush Limbaugh or Al Franken because they find them funny. In any case, fallacy or not, we need to be on guard against letting humor or ridicule distract or sway us. A joke is no substitute for an argument.

Update (1/30/2007): Reader David Short writes:

I thought I might point out that "ad festivus" is not a possible phrase. The preposition "ad" requires the accusative case after it here. "Festivus" is in the nominative, so it needs to become "festivum", just as "homo" becomes "hominem", "natura" becomes "naturam", "populus" becomes "populum", etc. In any case, "ad festivum" is no good anyway, nor "ad rusticum". "Ad ridiculum" ("to a joke") would be better. "Ad leporem" (from "lepos", meaning charm, grace, wit, or humour) might be better still.

I believe that appeal to humour is one of the most important and widely used fallacious arguments out there. I don't think it makes sense to say that it is not a fallacy because it is not a real argument, because the same applies to all fallacies―I don't consider that someone fighting a straw man or appealing to fear is actually making a real argument either.

If we absolutely have to have a Latin name for it, then "ad ridiculum" would be best. Given its relation to the English word "ridiculous", it would be easier to remember than any of the others.

Some traditional "fallacies", such as the various appeals to emotions like fear, often do not appear to be arguments. However, most fallacies are types of argument, in that their instances have identifiable premisses and conclusions, connected by indicator words such as "therefore" and "since". So, there needs to be a distinction between logical mistakes in argument and other problems.

I don't know why you would doubt that straw man arguments are, in fact, arguments. A straw man is a refutation, that is, an argument seeking to show that something is false. The fallacy in a straw man argument is that it shows the wrong thing to be false, that is, it refutes something other than what it was supposed to refute, or was claimed to refute.

Appeals to emotion or humor may not be arguments themselves, but distractions which either confuse or overwhelm us with feelings. So, by doubting whether "ad ridiculum" should be classified as a fallacy, I don't mean to question whether it can be a rhetorical dirty trick. It's just that it would be nice to have some other word than "fallacy" for this type of problem.


September 6th, 2006 (Permalink)

Check 'Em Out

September 5th, 2006 (Permalink)

Poll Watch

According to CNN:

Opposition among Americans to the war in Iraq has reached a new high, with only about a third of respondents saying they favor it, according to a poll released Monday. Just 35 percent of 1,033 adults polled say they favor the war in Iraq; 61 percent say they oppose it―the highest opposition noted in any CNN poll since the conflict began more than three years ago.

Reporters love to write stories about records, but critical readers should view them skeptically. As I've discussed here in the past, reports of record-setting prices or profits are often misleading because they don't take inflation into account. Similarly, I've criticized many stories about polls on the grounds that they ignore the margin of error. Major news outlets usually require their reporters to mention the margin of error, but they don't seem to insist that the reporters actually understand it, or incorporate their understanding into the story. Both of these phenomena came together to produce the above story.

The current CNN poll shows that 61% oppose the war in Iraq, which is apparently a record for that poll. However, it is only one percentage point higher than the same poll two weeks earlier. When the margin of error of plus or minus three percentage points is taken into account, what the poll shows is that opposition to the war in Iraq has not changed significantly in the previous two weeks. This is hardly surprising since nothing much has happened in those two weeks to cause any sudden changes in public opinion of the war. The "record" is well within the range of sampling error.


Resource: Jack Rosenthal, "Precisely False vs. Approximately Right: A Reader’s Guide to Polls", The Public Editor, 8/27/2006

Fallacy: Fake Precision

September 2nd, 2006 (Permalink)

Wang's Paradox

  1. One is a small number.
  2. If a number is small, then the result of adding one to it is also small.
  3. Therefore, a googol―one followed by a hundred zeroes―is a small number.

What's wrong with this argument? The first premiss is surely true. Moreover, one is too small an increase to turn a number from small to not small, so the second premiss also appears to be true.

Furthermore, the reasoning in the argument appears to be valid. Given that the second premiss is true, it follows that if one is small, then two is also small. But we know from the first premiss that one is small, so two is small, as well. Similarly, if two is small, then three is also. However, we just saw that two is small, so three is small. We can keep going on this way through a googol number of steps―it will take a long time, but we can do it if we're patient―until we reach the conclusion that a googol is small. But that is absurd! A googol is obviously not a small number, but a very large one. So, the conclusion is clearly false. What went wrong?

An argument such as this that seems sound, yet has an obviously false conclusion, is called a "paradox". Paradoxes are important in logic and philosophy because they reveal one of two things:

Thus, the study of paradoxes can help logicians identify logical fallacies, and the study of logical fallacies can help to resolve paradoxes. The argument above is a version of what's called "Wang's paradox" after the logician Hao Wang. What fallacy does it commit?

The word "small" is vague. Clearly, as the first premiss states, one is a small number. Just as clearly, contra the argument's conclusion, a googol is not a small number. However, in between one and a googol is a grey zone of borderline cases which are neither clearly small nor not. It is this fuzziness around the edges of smallness which makes it possible for the argument to start with uncontroversially true premisses and slip undetectably with each step a little further from the truth until it reaches a shockingly false conclusion.

If "small" had a precise meaning―say, any number less than 10―then the second premiss of the argument would be false, since 9 would be small but 9 + 1 would not be. The moral of the paradox is that we must be careful to reason only with precise concepts, or to keep a close watch on the vague ones lest they mislead us.

Resource: William Poundstone, Labyrinths of Reason: Paradox, Puzzles, and the Frailty of Knowledge (1988), pp. 94-95.

Fallacy: Vagueness

Previous Month | RSS/XML | Current | Next Month