A Memory Experiment That Would Make A Good Parlor Game

There are all kinds of ways to make people misremember things. But can you make a person misremember a specific word? This might be an interesting parlor game, or an interesting mass experiment.

We all know the game Taboo, in which people attempt to make their partner say a word or a phrase, while being unable to use the most common ways of describing that phrase. So a person might tell try to get their partners to say the word "unicorn," without using the words "horn, horse, myth, or forehead."

It seems there is a psychological procedure that works somewhat like Taboo, except instead of cooperation, the person listing the descriptive words is attempting manipulation. It's called the Deese–Roediger–McDermott Paradigm. It started as a way to instill false memories in the person hearing the list. Pick a word, and start listing words that evoke the sense of the word. So, if you were to pick "war," you might use a list of words that includes, "army, gun, soldier, politics, treaty, ceasefire," and so on. Read the list of words to a person, and then, after a time, have them try to repeat the list from memory. When they include the word that was the inspiration for the list — but never actually listed — it counts as a false memory, under the DRM Paradigm.

The basic format for this procedure has been around since 1959, and ever since then, people have been tinkering with it. The success of different lists varies wildly. A list containing, "nurse, sick, lawyer, medicine, health, hospital, dentist, physician, ill, patient, office, stethoscope, surgeon, clinic, cure," yields a 0.60 recall rate for the word "doctor." A list that contains, "queen, England, crown, prince, George, dictator, palace, throne, chess, rule, subjects, monarch, royal, leader, reign," gives you only about a 0.10 recall rate for "king." Exactly what is it that makes people falsely recall certain words but not others?

I'm not sure whether I'm calling for a wide scale experiment or a parlor game or both, but I think this might be a time for citizen science. People can start long-running games, trying to think up the perfect set of words for any one word, and then testing it out. They can register their lists, and their results online. We can see what combinations of words work best for what target words. We might even turn it into a contest. Or we can just spend a happy hour trying to twist our friends' minds.

[Via False Recall in the Deese-Roediger-McDermott Paradigm, On the prediction of occurrence of particular verbal intrusions in immediate recall]