Dr. Proulx and Dr. Heine described having 20 college students read an absurd short story based on “The Country Doctor,” by Franz Kafka. The doctor of the title has to make a house call on a boy with a terrible toothache. He makes the journey and finds that the boy has no teeth at all. The horses who have pulled his carriage begin to act up; the boy’s family becomes annoyed; then the doctor discovers the boy has teeth after all. And so on. The story is urgent, vivid and nonsensical — Kafkaesque.
After the story, the students studied a series of 45 strings of 6 to 9 letters, like “X, M, X, R, T, V.” They later took a test on the letter strings, choosing those they thought they had seen before from a list of 60 such strings. In fact the letters were related, in a very subtle way, with some more likely to appear before or after others.
The test is a standard measure of what researchers call implicit learning: knowledge gained without awareness. The students had no idea what patterns their brain was sensing or how well they were performing.
But perform they did. They chose about 30 percent more of the letter strings, and were almost twice as accurate in their choices, than a comparison group of 20 students who had read a different short story, a coherent one.“The fact that the group who read the absurd story identified more letter strings suggests that they were more motivated to look for patterns than the others,” Dr. Heine said. “And the fact that they were more accurate means, we think, that they’re forming new patterns they wouldn’t be able to form otherwise.”
In other words, when you start to break down people's sense that they understand what's going on, they respond by turning up the 'gain' on pattern detection. Similar things have been seen in previous studies, except in these studies the gain detection is turned up so high that people see things that aren't there at all.
For example, people who are made to feel like they are not in control tend to see patterns that aren't there. And people who are made to feel lonely are more likely to anthropomorphize (i.e. see pets and even gadgets as friends).
The interesting thing is that the NY Times ties this study in with an earlier one by Michael Inzlicht, on how 'error-related negativity' (ERN) predicts academic performance. ERN describes a brain signal that's triggered when you make a mistake.
The idea is that the bigger the ERN signal, the bigger the distress you get from things that don't make sense. Inzlicht showed that people with a big ERN response have better academic performance. They learn better.
The implication is that creating uncertainty increases the ERN, and so improves your ability to detect patterns and learn from mistakes.
Now, what the NY Times didn't pick up on is that Inzlicht published another study earlier this year (I blogged it here). This study showed that religious people have a low ERN.
So uncertainty increases, and religion reduces, ERN. It looks like a feedback mechanism to keep levels of ERN under control by reducing the level of ambiguity and uncertainty in the world (by 'explaining away' mysteries) - at the cost (perhaps) of failing to pick up on real, but obscure patterns.
Proulx T, & Heine SJ (2009). Connections from Kafka: exposure to meaning threats improves implicit learning of an artificial grammar. Psychological science : a journal of the American Psychological Society / APS, 20 (9), 1125-31 PMID: 19656338
This work by Tom Rees is licensed under a Creative Commons Attribution-Share Alike 2.0 UK: England & Wales License.