Being Wrong - Kathryn Schulz [39]
These confabulatory answers were the left brain’s solution to a strange problem. That side of the brain is heavily linguistic, and as such it is responsible for crafting our narratives about the world. The right brain, by contrast, is only minimally linguistic; it can understand commands and initiate actions, but it can’t generate explanations. In healthy human beings, this division of labor isn’t a problem, because information is constantly shuttled back and forth between the two hemispheres. In split-brain patients, however, the two sides have no way to communicate with each other. As a result, when Gazzaniga’s subjects were asked to account for their behavior, the right side of the brain (which had seen and responded to the commands) lacked the ability to explain what was going on, while the left side of the brain (which was able to generate explanations) lacked the requisite information. In other words, the left hemisphere literally had no idea why its own self was acting as it was. All it could do was theorize backwards from the subject’s behavior, and it proved extremely adept at doing so. With no apparent befuddlement, no noticeable time lag, and no appearance of doubt or intent to deceive, the left side of the brain consistently generated completely plausible—although, of course, completely wrong—explanations.
If confabulation occurred only as the result of brain damage or drastic surgical intervention, it would just be a freakish footnote to neuroscience.* In fact, though, it’s strikingly easy to get healthy people to confabulate. In 1977, the psychologists Richard Nisbett and Timothy Wilson set up shop in a department store in Michigan, where they asked people to compare what they claimed were four different varieties of pantyhose. In reality, all the hose were the same, but that didn’t prevent shoppers from showing a preference for one of them. Moreover, it didn’t stop them from explaining their preference, by claiming that (for instance) this color was just a little more appealing or that fabric was a little less scratchy.
In a sense, this is blind Hannah all over again. It’s weird enough that these shoppers chose between identical pantyhose in the first place, but it is even weirder that they generated explanations for those choices. After all, they could have just shrugged and declined to explain their decisions. We are expected to be able to justify our beliefs, but not so our taste. “I just like that one; I couldn’t tell you why” is a perfectly acceptable explanation for why we’re attracted to a particular pair of pantyhose (or a particular shade of blue, or a particular flavor of ice cream). In fact, it might be the only acceptable explanation: as we say, there’s no accounting for taste. Yet these shoppers insisted on providing accounts anyway. Since there were no differences among the pantyhose, these accounts couldn’t have been the real reasons behind the shoppers’ choices; they could only be post-hoc justifications. Their real motivations remain mysterious. The one factor the researchers could identify was the influence of position, since almost four out of five shoppers preferred the hose on the far right-hand side of the display over those on the far left-hand side. But of course, none of the shoppers explained their choice by reference to location. Instead, like split-brain patients, they confabulated explanations for a decision whose actual origins was buried in an unreachable part of the brain.
At first, this experiment seems to demonstrate a strange but basically benign quirk of human cognition: we like to explain things, even when the real explanation eludes us. But it has a sobering epilogue. When Nisbett and Wilson revealed the nature of the experiment to its unwitting subjects, many of them refused to believe that the pantyhose were identical. They argued that they could detect differences, and they stuck by their original preferences. Likewise, people who work with clinical confabulators report that the most striking