Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
SAF Archives  search ask the scientists in the classroom cool science
scientists from previous shows
cool careers in science
ask the scientists

Photo of Robert Stickgold What's in a Dream?

At Harvard University's sleep lab, Robert Stickgold's research is trying to establish how REM sleep and dreaming affect mental activities and the learning process. Robert answered viewers' questions about these topics as part of an Ask the Scientists panel following the premiere of the Scientific American Frontiers Special Pieces of Mind. Here are viewers' questions and her answers:

QIf you remember better if your REM sleep is uninterrupted would you remember better if you studied just before bedtime?

First of all, not all types of memory are enhanced by REM sleep, and we're not quite sure which kinds are. It seems that the more complex the information you're trying to learn, the more REM sleep is important. Just memorizing a list of words doesn't seem to be helped by REM sleep, but learning a foreign language (vocabulary and grammar) is helped. On the other hand, some kinds of very basic learning, known as procedural learning, might also be enhanced by REM sleep. Procedural learning has to do with learning how to do something, whether it's driving a car, playing tennis, or spotting unusual signals on a radar screen. Students studying for exams seem to have more REM sleep, which might be the real answer to your question.

But right before sleep? We really don't know exactly when the "REM window" is for different types of learning, and getting good REM sleep two nights later might be as important as getting it two hours later. At the risk of sounding like your mother, the best technique is to study when you're wide awake, alert, and able to focus your attention on the task, and then get a full night's sleep.

QHave there been any studies done on dreaming in animals other than humans--could you determine what they're dreaming about? For instance, my dog used to move his legs as if he were running and make short barks and grunts. And what about other primates--surely gorillas, chimpanzees and orangutans must dream. Do we know anything about it?

There's REM sleep, and then there's dreaming. The only way we can know for sure that anyone or any animal is dreaming is for them to tell us in words, so we can't really know that animals dream, or even that babies dream. But we do know that people dream, and that most of this dreaming occurs during REM sleep, and that you dream most of the time you're in REM sleep. So it's not unreasonable to guess that if animals have REM sleep they probably dream.

And they do have REM sleep -- all mammals and baby birds -- but not lower animals. So that's one reason to believe that they dream. The other reason is the one that we all know, whether we see our baby suddenly smile or start sucking while asleep, or watch our dog start barking quietly and "running" while fast asleep. I'm quite sure that all of these events have been shown to happen during REM sleep in infants and dogs.

One problem is that our voluntary muscles are all powerfully inhibited during REM sleep precisely to prevent us from acting out our dreams. (This is not true when we have those short, "hypnogogic" dreams when we first fall asleep, when we're not in REM sleep, and we have all those twitches that wake us back up.) Researchers have done experiments in cats where they have cut the nerve bundles that produce this inhibition and watched what happens. Sure enough, when the cats go into REM sleep that suddenly start leaping up, arching their backs, hissing, putting up their fur, etc., or else seem to be chasing and pouncing invisible mice.

So I guess I believe that the answer to your question is "Yes, animals do dream," but we're stuck with this sort of indirect evidence to support this belief.

QIf dreams are really caused by random images flashing into the brain, how do we explain recurring dreams?

Once the random images start flashing in our brains, our brains try frantically to make some kind of sense out of it. As it does this, it's going through your memories and associations, thoughts and ideas, trying to find something that will fit the sketchy outline that these random images are producing. It's sort of like it I suddenly said to you, "Tell me a story about a big bear and a little girl." You might make up some wild story, or you might end up telling me about Goldilocks and the Three Bears. And if I asked you the next night to tell me a story about an animal, you might come up with a totally new story, or you might retell the one from the night before. In both cases it has to do with how much the request (the random flashes or my question) limits the kind of responses that are allowed, and how strong some of the possible responses are in your mind/brain. If you absolutely love Goldilocks and the Three Bears (or if you've been slowly but surely traumatized over years and years by all the tests you've had to take), then that story (or that damn exam dream) might tend to pop up over and over again. So it's only the question (the initial stimuli produced during REM sleep) that's random. Not the answer (or the final dream constructed by your brain).

QHow do you expect to extend this research?

In the near future I'm hoping to look at the question of how dreaming relates to various memory systems within the brain. For example, it appears that we almost never replay actual waking events in our dreams. Let me give you an example. I might be driving home from work one day when someone runs a stop sign and we narrowly avert a disastrous collision. That night, I might dream about being at some amusement park, riding in a bumper car, getting hit by all those other little cars, and thinking, "This really isn't fun, why would anyone want to do this?" When I wake up in the next morning, I would probably immediately "realize" that this dream was sparked by my near-miss accident the day before, and all the memories of that near-miss would flood my mind. But I would never actually dream the memory of that near-miss; in my dream I wouldn't access and replay the memories that are sitting there in my brain somewhere. Why not?

Interestingly, the one time that people do seem to replay actual memories in their dreams is patients with post-traumatic stress disorder. And they seem unable to "work through" their traumas. Whether these two facts are related is an exciting question that relates to both how dreams are formed and how they function.

QWhat, if anything, influences the content of your dreams? i.e. stress factors? food you ate?

What doesn't influence your dreams? The problem in trying to answer this question is two-fold. First of all, nothing will reliably influence your dreams. When students watched a particularly gruesome movie before going to bed (things like new footage of mutilations and medical films of amputations), next to none of the subjects subsequently reported dreams that seemed in any way related to the movie, although all said that it had been very upsetting. The second problem, is that it's very difficult in most cases to argue convincingly that something was a definite influence on your dream. If you have subjects write down before going to bed what happened that day that was most important, these tend not to be what gets incorporated into your dreams. But nevertheless, it's clear that both recent and old events work their way into our dreams. We're doing a study right now on these questions, trying to get at the features of daytime events that end up in your dreams. Is it the emotions, characters, locations, actions, general

QWhy do you have rapid eye movement during a dream?

"Why" is a tricky question. If you mean "why" in terms of what causes them to occur, the answer appears to be that during REM sleep there is a REM control center in the brainstem which gives off periodic bursts of activity that spread over wide areas of the brain. These chaotically arising signals activate other midbrain structures that control eye movements. If the REM center on the left side of the brain happens to send out a burst, the eyes will move to the left; if the right side center fires, the eyes will move to the right.

If you're asking "why" meaning what's it for, the answer is less clear. One theory is that it's one of the way that the brainstem gives the forebrain the raw materials out of which to construct dreams; the forebrain recognizes that the eyes are moving and decides that something in their field of vision is moving, or that they're moving, and builds this information into the ongoing dream narrative. Interestingly, dreams turn out to be chock full of movement, much more than you would normally see in your waking life. Maybe this is why. But a second theory turns this idea on its head. This "scanning hypothesis" proposes that the eyes are actually following the action of the dream, much as your eyes follow the action on a movie screen. But the anatomy of the brain doesn't seem to encourage this explanation; there don't seem to be major pathways that could carry clear information about the dream imagery back down to the brainstem so that it could send signals up to the midbrain to make the appropriate movements.

QI noticed that you were looking at amplitude waveforms to discern brain wave activity. Could you utilize both magnitude and PHASE of the detected signal to discern additional information from the received signal? This could give you a better representation of activity within the brain for a 3d image.

I'm not sure what we would measure phase angle relative to. Recent work with 40 Hz brain waves has suggested that perception is in part mediated by phase locking of these signals in different parts of the brain, a phase lock that breaks down when the intactness of an image is no longer perceived. But with the waveforms we tend to look at, they are generally phase locks across the entire brain at all times.

QIf you can learn while you're sleeping, are there any experiments I could do at home to help prove this?

Not that we've been able to think of. It's really hard to figure out good tests for this. Most of the ones we've come up with require subjects to be in the sleep lab so we can either monitor the amount of REM and non-REM sleep that they get or to actually manipulate the amounts they get, for example by waking them up whenever they go into REM sleep so that they end up not getting any significant amount of REM during the night.

QWhy is it so hard to remember what we have dreamed about after we get up?

Although we don't have a firm answer to this one, it obviously has to do with how memories are formed, and the fact that they are formed differently when you're dreaming than when you're awake. There are, unfortunately, a couple of different theories. One says that because of the altered chemical environment in the brain during REM sleep, the brain is literally unable to form new memories. The only way to remember a dream is to wake up out of the dream while it's still fresh in your mind, and then rehearse it in the wake state. According to this theory, it's during the process of rehearsal that the actual memory is formed so that you can remember it later.

The second theory says that memory formation is fine during REM sleep, but that recall is very difficult. This theory is strongly supported by those times when something happens during the day, like a dog runs out in front of a car and almost gets hit, and you suddenly say, "Oh wow, I had a dream last night with a dog running around in it." The argument here is that the memory of the dream is there in your memory (when you recall the dog dream, you see the whole dream segment with clear visual imagery, recalling your feelings, what else was going on, etc.) but until the dog appeared the next day you had no associations that would allow you to locate and recall the memory of the dream. After all, if I try to remember where I left my car keys last night, I have a strategy for searching my memory -- what did I do after dinner, where did I go, when did I come home, what did I do when I came back into the house, etc. But if I want to recall my dreams from last night, where do I look for that memory? Unless we can hold onto the images in our mind when we first wake up, and then use then as a pathway to the memory of the rest of the dream, we seem to be out of luck.

QIs there really such a thing as a day-dream?

Sure. But they're not quite the same as night dreams. We tend to think of waking and sleeping as two states and we're always in one or the other. But as most of us know from watching the evening news, there's a state in between when we're not quite awake but not quite asleep either. In fact, there's a continuum of states, and day dreaming is one of them. We're clearly awake, but we've sort of cut our awareness of our environment, and have probably temporarily shifted our brain chemistry to a condition where its easier to create fantasies and even visualize them, although day dreams are almost never as vivid or "real" as REM dreams. Writers and painters probably go into similar "altered states" when they are in their creative periods; the "muses" that the Greeks used to talk about are probably brain chemicals and not the gods they thought they were.

QI was fascinated by your study of the effect of REM sleep deprivation on memory & learning. As a mother of three, It has always seemed that having children made me "stupid," so to speak; I sometimes found it hard to remember things I knew I should. Now I see that it could be contributed to a lack of REM sleep. Getting up every 2-3 hours. through the night is bound to interrupt some REM cycles. Has any research been done on interrupted sleep patterns like a new mother's? Over time, do the REM patterns adjust, does a mother go into REM sooner in order to compensate for the shortened sleep times?

What a wonderful idea! I don't know if anyone has done the study, but it would be fun to try. But I suspect that there's much more than just REM learning going on. I'm convinced that sleep deprivation enhances bonding; this is why people like to go on dates in the evening and stay up late. Something changes in how we feel about other people just as a function of this combination of circadian cycle and sleep deprivation. Besides this, there is a lot that you were learning during that period involving your new baby. We become so focused on them that maybe we just don't pay enough attention to other things in our lives to appear "smart" in those other spheres of our lives.


Scientific American Frontiers
Fall 1990 to Spring 2000
Sponsored by GTE Corporation,
now a part of Verizon Communications Inc.