We can quibble over the merits of being sensitive (touchy), confident (pushy), or thrifty (cheap). We can even speak ill of open-mindedness when taken by some moral relativists to its extreme. However, I would have thought empathy—the ability to understand another’s perspective—to be an unequivocal good, a virtue squeaky clean and beyond reproach.
But what if seeing ourselves in others encourages us to remember others as ourselves and thus take credit for things we didn’t do? This is the intriguing question examined in a 2010 study conducted by Sarah Barber and her colleagues at Stony Brook and Hokkaido universities. The researchers asked pairs of students to engage in collaborative storytelling: participants took turns contributing sentences to a story, adding anything they pleased so long as each sentence contained a single, critical word that appeared to both participants on a computer monitor. If the word fix appeared, for instance, a participant might begin, Once upon a time there was a girl who liked to fix xylophones. Then, a second word (refrigerator) would appear, and the other participant might say, She kept her tools in a box above the refrigerator. Compelling!
Once 80 sentences had been produced in this manner, the participants—whether their heroine was safely in bed or dangling precariously from a fourth-story ledge at the xylophone factory—were instructed to stop. They busied themselves with a filler task before returning, 15 minutes later, to the list of critical words. For each word, participants decided whether they’d contributed it to the story, whether their partners had, or whether it was one of 40 new words that had been mixed into the list.
Afterward, participants answered a number of questions designed to measure social intelligence, the tendency, as the researchers put it, “to anticipate a partner’s response across a broad range of circumstances and sources.” What did the researchers find? There was no relationship between individuals’ social IQ scores and their recognition memory—memory for whether a word was new or old. But participants’ ability to correctly identify the source of the word (i.e., who had added it to the story) was negatively correlated with social intelligence. That is, the more the participants agreed with statements like I can predict other people’s behavior, the less likely they were to correctly remember whether they, or their partners, contributed a word.
Psychologists have long understood that the more similar two possible sources of information are, the more apt we are to confuse them. This is why we’re much less likely to have a memory lapse like It was either George W. Bush or my three-year-old nephew who choked on a pretzel than one like It was either Sherlock Holmes or Hercule Poirot who solved A Study in Scarlet. But this finding suggests that actively anticipating a partner’s future behavior (in this case perhaps by generating potential sentences) is enough to muddle the distinction between other and self, leading to confusion about who, precisely, did what. Indeed, in a second version of the task, in which participants did not learn their partners’ critical words until after the sentence was supplied (and thus were unable to anticipate their partners’ contribution), the relationship between social intelligence and source attribution disappeared.
I like this study. I wish I’d run it myself (not least because transcribing hundreds of oddball stories sounds entertaining). But I’d find the results more compelling if the socially attuned storytellers had misattributed entire sentences to themselves, rather than the critical words, which, after all, were really just assigned by the experimenters. I hope a like-minded researcher does this experiment next. Not too like-minded, though. Otherwise, it’s going on my C.V.
