One of my favorite silly poems is “The Deacon’s Masterpiece,” by Oliver Wendell Holmes. It tells of a wonderful one-horse carriage, built in “such a logical way, it ran one hundred years to the day, and then of a sudden it” … well, it fell apart, not one part at a time but instantly, all at once.
What happened? The carriage’s designer had insisted that every part—leather, wood and iron—had to be perfect, and exactly as well-constructed as every other. As a result, when it broke down it did so altogether, with no part first, “just like bubbles do when they burst,” leaving just a heap of dust on the ground! The poem ends with the statement “Logic is logic, that’s all I say.”
As it happens, however, logic isn’t just logic, or at least, it isn’t always as straight-forward as one might think, specifically because even logic is filtered through that strange organ known as the human brain. And the evolutionary design features of the human brain may well hold the key to our penchant for logic as well as illogic. Following is a particularly revealing example, known as the Wason Test, which starts off being surprisingly tricky, and ends by being not only revealing but fun to contemplate.
Imagine that you are confronted with four cards. Each has a letter of the alphabet on one side and a number on the other. You are also told this rule: If there is a vowel on one side, there must be an even number on the other. Your job is to determine which (if any) of the cards must be turned over in order to determine whether the rule is being followed. However, you must only turn over those cards that require turning over. Let’s say that the four cards are as follows:
a b 2 3
Which ones should you turn over? (Remember, you want to assess this rule: If there is a vowel on one side, there must be an even number on the other.)
Most people realize that they have to inspect the other side of card “a” because the rule requires an even number on the other side. Similarly, most people realize that they don’t have to turn over “b,” since whether the other side is even or odd, the rule isn’t violated. However, a large proportion respond that the “2” should be inspected. They are wrong: The rule says that if one side is a vowel, the other must be an even number, but nothing about whether an even number must be accompanied by a vowel. (The side opposite an “2” could be a vowel or a consonant; either way, the rule is not violated.) Similarly, many people do not realize that the “3” must be inspected: if its flip-side is a vowel, then the rule is violated. So, the correct answer to the above Wason Test is that “a” and “3” should be turned over, but “b” and “2” should not be. Fewer than 20 percent of Harvard undergraduates get it right.
Next, consider this puzzle. You are a bartender at a night club where the legal drinking age is 21. Your job is to make sure that this rule is followed: People under 21 must not be drinking alcohol. Toward that end, you can ask individuals their age, or check what they are drinking, but you are required not to be any more intrusive than is absolutely necessary. You are confronted with four different situations, as shown below. In which case (if any) should you ask a patron his or her age, or find out what beverage is being consumed?
Under 21 Over 21 Drinking Water Drinking Beer
Nearly everyone finds this easy. You need to check the beverage of the person under 21; you needn’t check that of someone over 21; you needn’t check the age of someone drinking water, but you had better check the age of someone drinking beer. The point is that this problem set—which is almost always answered correctly—is logically identical to the earlier, the one that causes considerable head-scratching, not to mention incorrect answers.
Why is the second problem set so easy, and the first so difficult? This question has been intensively researched by the husband and wife team of Leda Cosmides and John Tooby, who conclude that the key isn’t logic itself—after all, the two problems are logically identical—but how they are positioned in a world of social and biological reality. Thus, whereas the first is a matter of pure logic, disconnected from the real world, the second plays into issues of truth-telling and the detection of social cheaters. The human mind, Cosmides and Tooby point out, is not adapted to solve rarified problems of logic, but is quite refined and powerful when it comes to dealing with matters of cheating and deception.
If you’re as intrigued and impressed by this “situational specificity” as I am, you can understand why findings of this sort make it increasingly likely that traditional cognitive psychology (which doesn’t take evolution into account) has been missing the boat: the mainstream “scientific” assumption that the human brain is a general-purpose information processing device is simply wrong. Perhaps 90% of previous research in the study of human memory involves nonsense syllables or geometric shapes precisely in an effort to avoid invoking the messy associations that human beings inevitably make with respect to anything real (such as cheaters, food, danger, sex, etc.). But we’re real, and so are the connections we make … and have difficulty making.
Logic is logic, that’s all I say. (Not.)
