Dreams Could Give Learning a Boost

Humans have long tried to figure out why we dream. In recent years, controversial research has suggested that we process our memories by dreaming. Now, a new study suggests that dreams also play a role in learning by processing what we've just experienced.

Researchers found that people who dreamed about a maze video game they'd just played did better the next time they tackled it, compared to those who didn't dream about it.

"There are parts of their brain which are actually replaying the memory of walking through the maze, and that will improve that memory and lead to a better performance," said study co-author Robert Stickgold, director of Harvard Medical School's Center for Sleep and Cognition.

Dreams have fascinated people for ages. "We started out a few thousand years ago thinking they were messages from God," Stickgold said. "Then Freud came along and said they're messages from our pernicious, immoral subconscious."

Scientists later thought dreams were "random firings of nerves in the brain stem," Stickgold said. Now, he said, "we're starting to say that at one level, as we've all sort of known all along, it's doing something with our memories."

In the new study, published online April 22 in the journal Current Biology, researchers asked 99 subjects to play a video game in which they had to find their way through a maze with the help of a three-dimensional depiction of it.

Then the participants either stayed awake for two hours or took a nap. They played the maze game again five hours later.

Four participants reported dreaming about the maze while they napped. They were among those who improved the most when they played the maze game for the second time, improving 10 times as much as others who napped.

So did the dreamers learn as they dreamed? It's not entirely clear how the dreams are connected to the experience of playing the maze game, but Stickgold thinks "the dream process reflects a type of underlying brain activity" that determines "what the learning experience means, not learning how to do it better."

There's another mystery: Why did so few people dream about the maze? Fewer than 10 percent of those who took naps did. By contrast, Stickgold said, about 86 percent of those who played an Alpine racer skiing game dreamed about it, he said.

Maybe the game "is not rich enough, not compelling enough," he said. Future research, he said, will try to figure out why that is.

For now, though, at least one brain researcher who is familiar with the study is unimpressed with the findings.

"There is no convincing evidence that sleep has any effects on consolidating memory," said Irwin Feinberg, a professor in residence who studies sleep at the University of California at Davis. Sleep clearly isn't necessary for people to remember things, he said.

But Stickgold said that's a misunderstanding of his research. "No one in the field suggests that you need to sleep to learn things or retain memories," he said.