What Do You Think?
Mark Twain once described religion as “a set of things which the average man thinks he believes.” When I first came across those words, I remember how intrigued I was at the idea of replacing “What do you believe?” with “What do you think you believe?”
People often lay claim to a belief that in fact they cannot hold. A belief needs to be defensible. Anyone who claims to believe everything in the Bible, for example, only thinks he believes these things. Since parts of the Bible flatly contradict other parts, it just isn’t possible to believe all of it. This is true of many beliefs that come in packages. It is sometimes said that an opinion is what someone says he has when he is still willing to argue his case, and to be persuaded that he is wrong; or that faith is, as H. L. Mencken put it, “an illogical belief in the occurrence of the improbable.” But in fact there’s a whole spectrum of belief-words, and there’s no way of ranging opinions, claims, notions, views, tenets, gut reactions, and viewpoints into any kind of foolproof order from the most airy of hunches to the most passionate of convictions. Nor does firmness of belief equate with probability of truth. In the words of Friedrich Nietzsche: “a casual stroll through the lunatic asylum shows that faith does not prove anything.”
To avoid drowning in a sea of warring definitions, I propose to label them all opinions (of varying strengths). This allows one to extend the line of inquiry by asking another question: Does a man have a right to his opinions? In a democracy, the answer yes is often taken for granted, and is defended with arguments about personal freedom. But the thoughtful answer is: “not necessarily.” The square root of forty-nine is not a matter of opinion. To put it more strongly, no one has the right to believe that it is eight, since this is an opinion that cannot be defended.
Readers of this newsletter will not need reminding that the fourth letter of CSICOP stands for the word “claims.” Indeed that concern is itself a claim. And sometimes skeptics with a misguided notion of tolerance assert that any claim whatever ought to be fully investigated. The purveyors of quack medicine have always been quick to pick up this point of view, and agree that (in the name of the open mind) medical researchers should drop whatever they are doing and spend their time endlessly testing and retesting pyramid power or faith healing or prayer. It is a common assumption that for any given subject, people must have an opinion; and furthermore that people know what their own opinion is. Neither of these assumptions need be true.
A survey organization once asked the simple question: “Have you ever heard of the Taft-Johnson-Pepper Bill on veterans’ housing?” Fifty-three percent of the respondents said yes. It can hardly be denied that these people had no right to their opinion, since the bill in question does not exist, and was simply invented for the purpose of the survey. Pollsters know that “the good respondent answers yes.” (People have learned that the answer no is often followed by disagreeable requests for justification.) When Britain’s Consumers Association asked people what they thought of the complementary medicine they had tried, 51 percent of them claimed it had improved them, and 31 percent claimed it had cured them. Notice that these people were not just saying they felt better: they were specifically claiming that the treatment was the cause of their improvement or cure - an opinion they had no right to (and that the Consumers Association had no right to ask them). It was a neat illustration of the philosopher Wittgenstein’s comment: “If there were a verb meaning believe falsely, it would not have any significant first person, present indicative.”
Tiny changes in wording can affect people’s stated opinions. One survey wanted to find out whether people favored limiting the president to one term in office. When asked if they favored doing this by “adding to” the Constitution, 50 percent said no. When asked if they favored doing it by “changing” the Constitution, 65 percent said no. Answers without responsibility also give notoriously poor quality opinions. If asked, “Would you like a free Skeptical Briefs with your copy of The Times?” most people would say yes. This would not show that most people like Skeptical Briefs. (Those who hated it could give the copy to someone else, or just throw it away.) The mention of even a small fee would change the answers drastically.
Respondents can’t even be relied upon with plain facts. In an Australian census, one-third of respondents claimed to have no ethnic origin at all. In a repeat American census, more than a third gave a different ethnic origin between one survey year and the next. When asked what they valued about a job, people ranked “good pay” first when asked an open-ended question, but they ranked it last when asked to choose from a list.
Lists themselves can affect answers. The last item on a list attracts about 10 percent more responses than when placed as the first item. A response can be affected by the response to the previous item. People claim to be much happier with their marriage if the response follows a previous question about happiness with things in general. In an agree-disagree question, about 22 percent will shift to “don’t know” if they are given that added option. People tend to agree with what they imagine the interviewer is thinking, even this involves contradicting themselves. They will agree that when it comes to crime, social conditions are more to blame than individuals; and later in the same interview they will agree that individuals are more to blame than social conditions. It is a reminder that “being positive” about something is often, in the words of Ambrose Bierce, “being mistaken at the top of one’s voice.”