Emeritus Professor of Psychology, London School of Economics; Visiting Professor of Philosophy, New College of the Humanities; Senior Member, Darwin College, Cambridge; Author, Soul Dust
Referential Opacity

It is commonly assumed that when people make a free choice in an election, the outcome will be what those on the winning side intended. But there are two factors known to cognitive science—but probably not known to politicians—which may well render this assumption false.

The first is the fact of “referential opacity,” as it applies to mental states. A peculiar characteristic of mental states—such as believing, wanting, remembering—is that they do not conform to Leibniz’s law. This law states that if two things, A and B, are identical, then in any true statement about A, you can replace A by B, and the new statement will also be true. So, if it’s true A weighs five kilos, it must be true B weighs five kilos, if it’s true A lives in Cambridge, it must be true B lives in Cambridge, and so on. The strange thing is however, that, when it comes to mental states, this substitution no longer works. Suppose, for example, the Duke of Clarence and Jack the Ripper were one and the same person. It could be true you believe the Duke of Clarence was Queen Victoria’s grandson, but not true you believe Jack the Ripper was her grandson.

It’s called referential opacity, because the identity of the referents in the two linked mental states is not transparent to the subject. It may seem and abstruse concept, but its implications are profound. For one thing, it sets clear limits to what people really intend by their words or actions, and therefore on their responsibility for the outcome. Take the case of Oedipus. While it’s true that Oedipus decided to marry Jocasta, it’s presumably not true he decided to marry his mother, even though Jocasta and his mother were identical. So Oedipus was wrong to blame himself. Or take the case of Einstein and the atomic bomb. While it’s true that Einstein was happy to discover that E=mc2, it’s far from true he was happy to discover the formula for making a nuclear weapon, even if E=mc2 is that formula. He said, “If I had known I would have been a clockmaker.” But he did not know.

Now, what about voters’ intentions in elections? Referential opacity can explain why there is so often a mismatch between what voters want and what they get. Take the German elections in 1933. While it’s true that 44% of voters wanted Hitler to become Chancellor, it’s clearly not true they wanted the man who would ruin Germany to become Chancellor, even though Hitler was that very man.

So, let’s turn to the second factor that can cause confusion about the real intentions of voters. This is the phenomenon of “choice blindness,” discovered by Lars Hall and Peter Johansson. In a classic experiment these researchers asked a male subject to choose which he liked better of two photos of young women. They then handed the chosen photo to the subject, and asked him to explain the reasons for his choice. But they had secretly switched the photos, so that the subject was actually given the photo he did not choose. Remarkably most subjects did not recognise the switch, and proceeded unperturbed to give reasons for the choice they had not made. Hall and Johansson conclude that people’s overriding need to maintain a consistent narrative can trump the memory of what has actually occurred.

Consider, then, what may happen when, in the context of an election, choice blindness combines with referential opacity. Suppose the majority of voters choose to elect Mr A, with never a thought to electing Mr B. But, after the election, it transpires that, unwittingly and unaccountably, they have in fact got Mr B in place of Mr A. Now people’s need to remain on plot and to make sense of this outcome leads them to rewrite history and persuade themselves it was Mr B they wanted all along.

I’m not saying the vaunted “democratic choice,” the “will of the people,” is a mirage. I am saying it would be as well if the wider public knew that scientists say it should be taken with a pinch of salt.