The Stuff of Dreams

The old Freudian notion that dreams are a window into the unconscious--revealing desires and fears that we'd rather not consciously face but that a therapist can help us excavate and confront--makes many neuroscientists today raise their eyebrows. That skepticism was clear in NOVA's program "What Are Dreams?" as well as in what sleep-and-dream researcher Bob Stickgold had to say to the flood of viewers who e-mailed questions after the show aired. (The gist: "Freud was probably 50 percent right and 100 percent wrong!")

Stickgold and his colleagues continue to rack up evidence that dreams are nonetheless a reflection of critical processes at work in the brain as we sleep, neurological housekeeping that helps consolidate and organize our memories and thereby enhances learning. A study in the April 22 issue of Current Biology led by Erin Wamsley, a postdoc in Stickgold's lab at Harvard Medical School, offers an intriguing case.

Wamsley and her team taught 99 Harvard undergrad volunteers how to make their way through a computer maze, represented by the schematic below.

Maze in Memory Study.jpg
(Courtesy Erin Wamsley)

The students had to remember the location of a particular object (the checkered cone) and then find the cone as quickly as possible when they were plunked down at different spots within the maze. After a first round of testing their navigation skills, half the group was led into a sleep lab for a 90-minute afternoon nap. The other half stayed awake, quietly watching videos.
Five hours after the initial testing, the students tried the maze again. Those who had stayed awake fared no better, even those who had mused about the maze in the interim. Those who had slept but didn't report dreaming about the maze made slight improvements. The scores of four nappers who had dreamed about the maze, however, skyrocketed. They became maze-navigating experts, performing 10 times better than the others who had slept.

When I spoke to her this week about the results, Wamsley stressed that her study doesn't show that dreaming causes the enhanced performance. Rather, she says, "Dreaming is a window into the memory-consolidation process. Whatever their brains are doing to work on these problems, we can see a reflection of that in their dreams." No one knows if dreaming causes learning or learning causes dreaming.  

The researchers can see, though, that the dreams don't appear to be exact replays of the experience of navigating the maze. (One student described his as a mash-up of images of the computer maze together with scenes from a bat cave he had once explored.) Dreams, at least some of them, seem to be by-products of the brain trying to put new memories into the context of what we already know.

There are, of course, different types of memories--the maze-navigating memories Wamsley looked at may involve different brain processes than those for reciting lines of Shakespeare or reminiscing at a high-school reunion. And we dream during different stages of sleep, both REM and non-REM. Wamsley's study focused on non-REM sleep, the stuff of 90-minute naps. She notes that other types of memories and learning--honing motor skills like salsa dancing, for instance--may rely more on REM sleep, which we generally only experience overnight after several hours of sleep.

So if dreaming about a task is associated with getting better at it, can we somehow direct our dreams to achieve the same effect? "That's what everybody wants to know," says Wamsley, but she's not certain it's possible. While people might be able to use imagery-rehearsal therapy before sleep to change the nature of their dreams (and banish recurrent nightmares), it's unclear whether this would trigger memory-consolidation.

There is also the fascinating phenomenon of lucid dreaming--a state in which people are aware that they are dreaming and can steer the course of their dreams. (I've done it myself, and like many lucid dreamers, I chose to go flying like Superwoman--fun!) But lucid dreaming seems to be a state of mind somewhere between REM sleep and wakefulness, and whether it allows dreamers to select and control which memories are processed or actually impairs the processing of memories is a big unknown. It's one of the many dream/memory connection questions researchers like Erin Wamsley and Bob Stickgold are continuing to probe. Stickgold says, "Check back in five years!"

User Comments:

I have a very poor memory. Perhaps the answers you find about dreams-memories-learning would help many like me.

blog comments powered by Disqus