There are two holy grails that are frequently studied but rarely achieved. One that is shared by grade school students, teachers, parents, and probably everyone else too is the ability to learn while you sleep. I have seen studies that looked at playing educational audio during sleep in the hopes that some of it will sink it. These never seem to work.
Another holy grail is the reduction of prejudice. All kinds of biases continue to dog our thought processes and behaviors. Reducing these could improve decision making as well as societal and community relations. Even promote world peace when the source of conflict has religious, ethnic, or racial components.
So this study got a great deal of coverage as soon as it hit the social media radar. It was covered in traditional science publications like Discover Magazine, long read sites such as Big Think, and shared on every social network I subscribe to (which is quite a few).
We wondered whether sleep could play a role in undoing implicit social biases. These are the learned negative associations we make through repeat exposure – things like stereotypes about women not being good at science or biases against black people. Research has shown that training can help people learn to counter biases, lessening our knee-jerk prejudices, many of which can operate without our notice.
It has long been known that much of long term memory is established while we sleep. During the day, a wide variety of mental models are created, including conceptual, episodic, and procedural categories. But it isn’t until the brain has some free time that the hippocampus and prefrontal cortex create the wiring changes that turn the mental models into enduring long term memories. This consolidation process happens during sleep.
But not everything gets consolidated. Only those experiences (whether episodic or ideational) that received strong enough activation during the initial experience to have some residual energy during sleep. This is why midday naps are good for learning. They help us consolidate our morning experiences before they fade away.
The Northwestern study that made such waves manipulated the consolidation process. Ever since Pavlov’s dog, we have known that if we pair a tone with an experience, just the tone can activate the expectation of the experience. The researchers wondered if this could be used to influence consolidation. Playing audio of real content while we are sleeping has never been shown to increase its activation during sleep and lead to consolidation. But what if content is learned while we are awake and paired with a very simple tone? Could that tone break through to our sleeping brain and increase the consolidation of the paired experience?
The research suggests that it can. Using the Implicit Association Test (IAT) as a baseline, the researchers measured racial and gender stereotypes. There are some validity issues with the IAT, but as a general measure of biased thinking it is a reasonable estimate to use as a baseline. Then they gave participants a simple but counter-stereotype association task. For example they showed them examples of famous women leaders in science, which runs counter to the gender stereotype. They paired that learned association with a tone.
Then while the participants took a post-learning nap, they played the tone. The idea was that the tone would activate the paired association during the consolidation process and increase the chance that it would be included and converted to long term memory. While playing the association during sleep is too complex, perhaps a simple tone that is paired with the association could reactivate it. And it turns out, it can. They reran the IAT on participants after the nap and there was a decrease in the gender and racial stereotypes.
The study just looked at short term changes. It is unclear whether participants would exhibit any long term changes in implicit bias – whether it would really be enough of a change in long term memory to alter behavior in a meaningful way. But whenever you are studying a holy grail, even a small change is positive. And when you combine two holy grails, even better. But before jumping on the bandwagon (as much of the media seems to be doing), I think we need to see the results of a more longitudinal study first.
One suggestion I saw was to create an app that runs every night. While the user sleeps, the tone would play periodically throughout the night to make sure that it hit the consolidation period at least once. So even if the effect is limited to short term, the app could make it long term. Perhaps the user would need to reinforce the pairing of the tone to the counter-stereotype learning every once in a while as well. But an app would be perfect for that as well. As long as the user was committed to it.
A more black hat version of the finding also made the rounds. Marketers could use the same phenomenon to create a paired association between their brand and some positive images or attributes. Then a nightly tone through the company’s app (perhaps distributed using a loyalty program or discounts) could artificially inflate the user’s impression of the brand. That would be evil, but we know it works. Think of the study that showed when a liquor store played French music, sales of French wine increased but when they played German music, sales of German wine increased. This is really the same effect, but applied to long term memory consolidation rather than working memory priming.
We are posting this on Thursday to give you the weekend to consider what you really think about this finding. Replication is probably needed before anything conclusive can be even considered. But does it give you hope? Do you think it is real? Will it be coopted by marketers as I fear?
Please let us know in the comments.
Image Credit: Gisela Giardino