The Elephant in the RoomPrint
By Jessica Love
September 29, 2011
Not long ago, on the phone with my brother, I suggested he squeeze a holiday dinner into his busy schedule. “My mom would love to see you,” I said.
“Mmmm,” Jimmy agreed. After a short pause, he reminded me, “You know, she’s my mom, too.” Then, after laughing with what seemed like unnecessary gusto, he said, “Don’t you, like, study pronouns?”
Of course, I needed no reminding that Jimmy and I shared a mother. Barring some unforeseeable tragedy, I am unlikely to forget all those times I made him cry, or clamped a barrette in his hair and introduced him to strangers as my sister Jimina. And yet, such cherished memories aside, I’d messed up. I’d mentioned our mother to him the way I’d mention her to a friend, a colleague, a cousin—to just about anyone who wasn’t him. As linguists would say, I’d failed to appropriately incorporate common ground—shared knowledge, the information I knew he knew I knew—into our conversation.
Researchers have spent a good bit of time trying to figure out just how common ground works—how and how well we’re able to keep track of it. Sid Horton (now at Northwestern) and Richard Gerrig of SUNY-Stony Brook have a theory I find compelling. Rather than positing any particular tracking mechanism—a mental diary or checklist, perhaps—they suggest something much simpler. Encountering someone, they argue, becomes a retrieval cue of sorts for everything we have experienced with that person. In short, if we’ve spent an afternoon together watching elephants at the zoo, we’ll spend future meetings with all things pachydermal more accessible in memory.
In order to test this, Horton instructed undergraduates to engage in a series of tasks. First, participants were asked to generate a list of English words. A research assistant sitting nearby–we’ll call her Shelby–read aloud a hint (e.g., type of vegetable) and then a fragment appeared on a computer screen (e.g., C _ _ R _T). Once the word carrot was generated (if the participant couldn’t come up with it, Shelby would tell him the answer), the next hint was given. Halfway through the list of fragments, Shelby left the testing room and a second research assistant, whom we’ll call Tony, took her place.
After the fragment completion task, participants were led into another room. Here, they saw pictures of objects on a computer screen and were instructed to name them as quickly as possible into a microphone. Again, half of the trials occurred in the presence of Shelby, and the other half occurred with Tony in the room. Importantly, some of the object names were words generated during the earlier task (e.g., a picture of a carrot was shown). Horton reasoned that participants would be faster to name an object with Shelby in the room if they’d initially generated the word with Shelby, and with Tony in the room if they’d initially generated the word with Tony—indeed, as it turns out, 86 milliseconds faster. Believe it or not, this is a sizable priming effect (I’ve danced a jig over significant priming effects half that size).
Here’s the kicker. Before leaving the experiment, participants saw one last list of words. For those words they remembered generating, they then had to decide whether Shelby or Tony had given them the hint. Although participants were overall quite accurate (85 percent), there was no relationship between the accuracy for any given item and the size of the priming effect. That is, participants were still faster to say carrot in Shelby’s presence even if they couldn’t explicitly remember that Shelby had been the one present when carrot was generated.
Although the occasional muck-up is inevitable (even, Jimmy, among those of us who study pronouns), the people in our lives do seem to help trigger shared information. This has its downsides, which should come as no surprise to anyone who has managed to suppress the details of an awkward encounter with a professor, only to have them come rushing back years later when trapped in an elevator with said professor. Mostly, though, through one conversation after another, common ground keeps us from embarrassing ourselves in the first place.
Jessica Love is a contributing editor of the SCHOLAR. She holds a doctorate in cognitive psychology and edits Kellogg Insight at Northwestern University.