The Frontal Cortex

Dreaming, Smelling and Memory

After Freud lost his scientific credibility, psychology became very dismissive of dreams. The leading scientific theory held that dreams consisted of mental detritus, the scraps and fragments of memories that your brain didn’t want to remember. While Freud mined our nighttime thoughts for hidden meanings – they were symbol laden narratives of wish-fulfillment – this modern theory held that our dreams were entirely meaningless, a montage of hallucinations. They were the result of our hippocampus taking out the trash.

But dreams have been slowly making a comeback. There’s now a large body of evidence demonstrating that certain parts of the sleep cycle are an essential part of the learning process. While you are sleeping, your brain is carefully consolidating your knowledge, figuring out what you should remember and what you should forget.

That’s the larger theoretical context for this latest study of dreaming, which demonstrated that, even while we are sleeping, sensory stimuli (especially smelly stimuli) can affect what our brain chooses to remember.

Scientists studying how sleep affects memory have found that the whiff of a familiar scent can help a slumbering brain better remember things that it learned the evening before. The smell of roses — delivered to people’s nostrils as they studied and, later, as they slept — improved their performance on a memory test by about 13 percent.

In the study, neuroscientists from two German institutions, the University of L├╝beck and the University Medical Center Hamburg-Eppendorf, had groups of medical students play a version of concentration, memorizing the location of card pairs on a computer screen. Upon learning the location of each pair, the students received a burst of rose scent in their noses through masks they wore. The researchers delivered the fragrance in bursts because the brain quickly adjusts to strong smells in the air and begins to ignore them.

The students went to sleep about a half-hour later, with electrodes on their heads tracking the depth of their slumber. Neuroscientists divide sleep into stages, including deep (or slow wave) sleep and the shallow, dream-rich state called rapid eye movement (or REM) sleep.

The brain is thought to process newly acquired facts, figures and locations most efficiently in deep sleep. This restful state usually descends within the first 20 minutes or so after head meets pillow and may last an hour or longer, then recur once or more later in the night. The researchers delivered pulses of rose bouquet during this slow-wave state; the odor did not interrupt sleep, and the students said they had no memory of it.

But their brains noticed, and retained an almost perfect memory of card locations. The students scored an average of 97 percent on the card game, compared with 86 percent when they played the game and slept without being perfumed by nighttime neuroscience fairies.

The students did not get the same boost when they received bursts of the fragrance just before sleep or in REM sleep rather than in deep slumber, and their improvements were not due to practice, the study found.

Just think about how much more Proust would have remembered if he’d smelled the madeleine in his sleep.


  1. #1 Ryan Fox
    March 9, 2007

    What happens if you use the same smell for two or more different memory-based exercises?

    For instance, as an engineering student, I get to experience stuffy air that smells like sweaty men up to 6 times per day in my classes…

  2. #2 Simon
    March 9, 2007

    The article says that the fragrance does not have any affect when administered in REM sleep, the stage of sleep with which dreaming is usually associated. Dreams do happen in deep sleep, but they are typically brief and, subjectively, less vivid that dreams in REM.

  3. #3 Jonah
    March 10, 2007

    That’s a good point, Simon, and I should have made it clear in the post. Matthew Wilson, who is a pioneer in this sort of dream research, compares REM dreams to movies and dreams during deep sleep to music videos. They are short, slightly incoherent, and usually lack a strong narrative. But they are still dreams. This is why Freud was wrong about the interpretation of dreams – they don’t reveal our inner psyche – but right to point that our dreams perform an important function.

  4. #4 Bryan
    March 10, 2007

    I’m not actually sure that this implies that dreams actually do perform an important function, as you’ve suggested, Jonah. Even if it were found that some other stimuli presented during REM sleep were able to improve memory consolidation, it does not necessary follow that the dreams, per se, were responsible for that. Dreams, like so many other human quirks, may simply be an interesting by-product of the specific pattern of neural activation during sleep.

    That being said, it still remains to be seen how much memory consolidation normally occurs for the average person, sleeping in their bedroom with their normal assortment of stimuli. Or, for those of us poor buggers still in school who live in cheap industrial centres, what effect noxious odours have on this process. Or other types of stimuli, particularly auditory stimuli. One of my colleagues has suggested that background noise during sleep (he thinks it’s REM, but it could be SWS) can also improve consolidation. This is one of those neat areas that’s still pretty wide open, and will be exciting to see what comes next.

  5. #5 limewire indir
    March 14, 2009

    of sleep with which dreaming is usually associated. Dreams do happen in deep sleep, but they are typically brief and, subjectively, less vivid that dreams in REM.

New comments have been disabled.