Traffic Dead Rise Slowly
Yeah, they're dead! They're all messed up.
Source: Richard Lederer, Anguished English (1989), p. 87
New Book: Thinking, Fast and Slow
Psychologist and Nobel Prize winner Daniel Kahneman has a new book out, and Slate has sort of a review of it by Daniel Engber. This is likely to be an important book, if Engber is right in describing it as "a compendium of [Kahneman's] thought and work". For the past four decades, Kahneman has been in the forefront of psychological research into how people make mistakes in reasoning. Unfortunately, logicians have been slow to pay attention to developments in the psychology of reasoning, and those who work under the rubric of "critical thinking" seem not to have been much quicker, as far as I can tell.
According to Engber:
…Kahneman designates no fewer than three biases (confirmation, hindsight, outcome), 12 effects (halo, framing, Florida, Lady Macbeth, etc.), four fallacies (sunk-cost, narrative, planning, conjunction), six illusions (focusing, control, Moses, validity, skill, truth), two neglects (denominator, duration) and three heuristics (mood, affect, availability).
Most of these are new to me, at least under these names, and as far as I can tell there is no overlap with traditional logical fallacies. Of course, confirmation and hindsight bias are familiar friends that I've discussed here often, but I've never heard of outcome bias. Three of the four effects that Engber lists are unfamiliar to me under those names. However, I've apparently heard of the halo effect before, because Engber claims that it's mentioned in a book that I've read, but I don't remember it!
Of the fallacies, the conjunction fallacy has a Fallacy Files entry of its own, and the sunk-cost fallacy is often discussed in economics, but I've never heard of the other two. The six illusions are new to me, except that Engber mentions the last one in the review, where he claims that repetition, legibility, and simple language add up to the "illusion of truth": "…a feeling of 'cognitive ease' that lulls our vigilant, more rational selves into a stupor." The two neglects are new, as well as two of the three heuristics―availability I've discussed here before.
I wonder how well Kahneman defines and distinguishes between these six categories, for instance, what's the difference between a bias and a fallacy? Between a bias and a heuristic? Between a fallacy and an illusion? Does Kahneman define a "fallacy" as a logical mistake, or in some psychological sense? For instance, the sunk cost fallacy has always seemed to me to be a specifically economic mistake and, therefore, not a logical fallacy.
Engber seems concerned in his review with what is to be done about such mistakes, because Kahneman is pessimistic about what people can do on their own to improve:
Again and again [Kahneman] reminds us that having the means to describe your own bias won't do much to help you overcome it. If we want to enforce rational behavior in society, he argues, then we all need to cooperate. Since it's easier to recognize someone else's errors than our own, we should all be harassing our friends about their poor judgments and making fun of their mistakes. Kahneman thinks we'd be better off in a society of inveterate nags who spout off at the water-cooler like overzealous subscribers to Psychology Today. … This imaginary world of psycho-gossip and thought correction sounds like a very annoying place. And while Kahneman's book offers some clear and engaging examples of how our minds work―or don't work―it's never clear whether the propagation of his catchphrases would really improve our lives. Even if organizations and governments can benefit from a rich language of cognitive bias, what would it mean for individuals? Do new ways of talking lead us to make better judgments from one day to the next?
I've often wondered about these things, as well, or at least the related issue of whether studying logic actually improves one's reasoning. However, identifying a disease and curing it are different problems, but the former is a necessary condition for the latter. Does studying fallacies improve reasoning? I'm not absolutely sure that it does, but I am sure that the question is an empirical one, and that we won't know the answer until psychologists like Kahneman put the question to the test. If our current teaching methods are ineffective, recognizing that fact is a necessary step in coming up with more effective ones. Also, Kahneman's work may help here, since the use of repetition, legibility, and simple language would give in this case not an illusion of truth, but the reality.
I'm skeptical that nagging people about their mistakes is a good idea, let alone an effective way of improving their performance. Accusing someone of committing a logical mistake by naming the fallacy strikes me as counter-productive. Instead, I think it's better to explain what's wrong with an argument. If you can't do this, you don't understand the fallacy well enough to be levelling the charge.
Engber also raises another question that concerns me:
Is there a point at which we'll have reached a state of overdiagnosis, where these self-help catchphrases have become so plentiful and diverse that we can no longer remember what they mean? … Eventually we'll be so inundated with "effects" that the word effect will lose its effect. Maybe that's already happened.
As I mentioned, "the halo effect" is a case in point, since I can't remember what it refers to despite having read a book that discusses it. I have no doubt that there are already too many named logical fallacies to expect the average person to learn and remember, but how many is the right number I don't know. Again, this is pedagogical question for psychologists to answer.
No doubt there are too many species of butterflies for most people to learn and remember; there may even be too many for any single specialist to know. However, the question for the lepidopterist is not how many can he or she remember, but how many there are. I'll leave the appropriate analogy as an exercise for the reader.
Source: Daniel Engber, "The Effect Effect", Slate, 10/26/2011
There was another debate between the Republican candidates for President a couple of days ago, this one in Las Vegas and moderated by CNN anchor Anderson Cooper. They seem to have one of these every week. Here's the very first question of the debate, which came from an audience member, and part of Congresswoman Michele Bachmann's answer:
Question: This is for all candidates. What's your position on replacing the federal income tax with a federal sales tax?
As a reason to oppose Cain's proposal, Bachmann invokes a slippery slope from a 9% national sales tax to a 90% one. But just how slippery is that slope? To support the slipperiness of the slope, Bachmann gives the example of the income tax going from 7% to 70%. Let's examine that example more closely.
Let's assume that Bachmann's numbers are correct―after all, this is a logic check, not a fact check. Why did she choose to compare the top income tax rate when the tax was adopted to that in 1980, that is, 31 years ago? Why didn't she compare the original top rate to the current one? I don't know, but the current highest marginal rate is 35%, according to Wolfram Alpha. This undercuts Bachmann's argument since, in the past three decades, the top rate has slid backwards to half of what it was.
So, what's to stop Cain's 9% national sales tax from ballooning to a 90% one? Presumably, the same thing that has prevented the income tax from doing so. Moreover, what's to stop Congress from adopting a national sales tax now? What has stopped it before now? Presumably, the usual mechanisms of democratic politics.
Fallacy: Slippery Slope
If you've visited this site before, you'll notice a change to the navigation frame on the left. The change is partly cosmetic, but primarily to simplify navigating around the site. The main menu is now contained in a table at the top of the navigation frame, followed by the complete alphabetical list of fallacies. The weblog also now appears on the front page where the main menu used to be. As with any substantial change, there are probably some remaining bugs, for which I apologize. If you notice a bug, such as a broken link, or have comments or suggestions about the site redesign, please let me know.
The Riddle of the Sheep
Why do white sheep eat more than black ones? (Traditional riddle.)
Another debate amongst the Republican candidates for President was held a couple of days ago. One of the questioners, Washington Post reporter Karen Tumulty, asked Texas Governor Rick Perry the following questions:
Governor Perry, over the last 30 years, the income of the wealthiest 1 percent of Americans has grown by more than 300 percent, and yet we have more people living in poverty in this country than at any time in the last 50 years. Is this acceptable? And what would you do to close that gap?
I don't know where Tumulty got her first claim, so let's just assume that it's true. However, her second may set off alarm bells for anyone old enough to remember a significant fraction of the last fifty years. Admittedly, the economy is currently in bad shape, and the unemployment rate is high, so it's not surprising that more people are in poverty now than a few years ago. But can poverty really be worse now than at anytime in the last half-century?
At first glance Tumulty's claim seems alarming, but notice her exact words: "we have more people living in poverty in this country than at any time in the last 50 years". So, all that she's claiming is that the absolute number of poor people is greater now than it was fifty years ago. However, the absolute number of people in the country is greater today than it was fifty years ago. According to Wolfram Alpha, the U.S. population was 184 million in 1961, whereas it is currently 309 million. Moreover, the population has increased steadily in the intervening fifty years. Thus, even if the percentage of the population in poverty stayed the same, the absolute number of poor people today would be greater than at any time in the past fifty years.
In fact, I think that I've found the ultimate source for Tumulty's claim in a press release issued by the Census Bureau a little over a year ago, which states the following: "The number of people in poverty in 2009 is the largest number in the 51 years for which poverty estimates are available." But it also states: "The poverty rate in 2009 was the highest since 1994, but was 8.1 percentage points lower than the poverty rate in 1959, the first year for which poverty estimates are available." So, the poverty rate has in fact declined considerably in the past fifty years, though the absolute number of poor people has increased because the population has increased.
Thus, what Tumulty said was literally true, but misleading. I suspect that almost everyone who hears it will assume that it means that poverty has increased in the U.S. in the last fifty years, rather than decreased. It's also a good illustration of why it is important to use proportions, rather than absolute numbers, when comparing changes over time when the size of the population is changing. Unless the population size is fixed, comparing absolute numbers can give a misleading picture.
Answer to the Riddle: White sheep eat more than black sheep because there are more of them.
New Book: Denying Science
A new book by John Grant sounds promising, at least based on its title: Denying Science: Conspiracy Theories, Media Distortions, and the War Against Reality. Grant is also the author of some other books on related topics, including Discarded Science, Corrupted Science, and Bogus Science. Unfortunately, I haven't read any of these and thus don't know what to expect from the new one.