Section 1 - Sleep and Dreaming



1.1 Do We All Dream?

 Research has shown1 that REM (rapid eye movement) sleep is prominent in humans and other terrestrial, placental mammals, with minor exceptions.38,79  It is known that humans dream during the REM stage because when awakened during REM sleep, laboratory subjects reported dreams or dream-like experiences an average of 82% of the time.39 Domhoff77 states that there are four conditions required for dreaming: 1) an intact fully mature neural network for dreaming; 2) a mechanism for activating the dream; 3) exclusion of external stimulus; and 4) loss of self control or shut down of the “cognitive system of self.”   It is uncertain what animals experience during this stage, since the “network for dreaming” in the animal brain is uncertain.

Early thought was that dreaming was initiated by REM activation.  However, the dream or dreamlike state is now known to occur also at the onset of sleep, during NREM sleep and even in some waking states.  Dreaming occurs to a lesser degree between REM cycles, with the average rate of dream recall being only about 42%.  There is also a marked difference in the dreams recalled from these two states. According to the research findings compiled by Hobson,39 reports from REM states are typically longer, more vivid, more animated, more emotionally charged, have more dream elements in them and are less like waking life events than NREM reports. NREM reports are more thought-like, and contain more representations of actual waking life events than do REM reports.  Krippner et. al.67 reports on a study by Hobson (performed on 146 dream reports by 73 subjects) using an equal number of REM and NREM dreams.  Hobson applied a standardized “Bizarreness Scale” to the dreams. The results indicated that about 70% of the REM dreams were reported as bizarre, compared to 20% of the NREM dreams.

 The time spent in dream sleep varies with age.  Adults spend about 25% of their sleep time in REM, children about 50%, and in premature infants it has been reported to be as high as 70% to 80%.37  In older persons, REM sleep may reduce to about 15%.  Art Funkhouser69 indicates that dreaming continues as people age, but the themes they dream about change, as do their daily concerns. It seems there are fewer nightmares, less frequent aggressive dreams, shorter average length of dreams and less frequent dream recall.  While shorter dreams and lower dream recall can be partly explained by the fact that memory no longer functions as well as it used to, research into aging and dream recall has shown that there are other factors involved since the largest drop-off in dream recall occurs at a relatively young age, around 26 (especially among males).70  As we age, the content may eventually evolve into concerns about death.  Research on this subject can be found in a book by Mary-Louise von Franz.71


1.2 Sleep Cycles

 Figure 1 illustrates the cycles of sleep in a normal human that occur every night.1

The horizontal scale is sleep time in hours, and the vertical the stages of sleep.  In simple terms, it shows that we cycle between deep NREM sleep and shallow dreaming REM sleep multiple times per evening.  These cycles continue throughout the night, in approximate 90-minute periods.  We typically go through four to six dream periods in an eight-hour night, which might be surprising since we are often unable to recall even one of these dreams.


Figure 1

Sleep and Dream Stages


Figure 1 illustrates a night of sleep.  Dream sleep or REM sleep is also known as D (for desynchronized or dreaming), stage 1 sleep, or “slow wave” sleep.  Non-REM or NREM sleep is also known as S sleep (for synchronized sleep) or stage 4 sleep.  Often at sleep onset we awake with a jerk, sometimes recalling some dream-like images.  This is known as a period of hypnogogic sleep, or light sleep where we have entered REM sleep and re-awakened.  Normally at sleep onset there is a rapid movement from stage 1 sleep to stage 4, i.e., we go right into the deep NREM sleep.  Pulse, respiration rate and blood pressure are lowered and no eye, facial or body movements are noted.  The brain waves are more regular and lower frequency (1 to 4 cycles per second).  Muscles relax, although whole body jerks may be observed.  In this deep sleep stage it becomes more difficult to arouse the person. If awakened, dream reports, if any, tend to be more thought-like and devoid of imagery.

 About 90 minutes later, the sleeper will begin REM sleep and vivid dreaming.  As we continue through the night, we enter deep S sleep less often, and after about 6 hours remain between stages 1 and 2.  The REM sleep stage is characterized by: eye movement, small movement in the muscles of the face, faster and more irregular pulse and respiration, and higher blood pressure. An EEG reading, called a PGO spike, often marks the onset of REM and many of these spikes appear during REM.  Resting muscle potential is almost non-existent (we can’t move).  It is easier to arouse someone from this stage. When aroused, dream reports are full of vivid imagery and content, more so at the end of the sleep period than at the beginning.  

 In addition to the differences in dream reports, there are differences in brain activity observed between NREM and REM sleep.  The medulla of the brain appears to regulate NREM sleep, whereas it is the pontine brain stem that regulates REM.1  Single neurons in the brain have been seen to reduce their activity during NREM sleep and then increase during REM to a level as high or higher than when awake.  During REM, a section of the forebrain is in a state similar to that of alert waking, and high levels of activity are found in the visual associative cortex and parts of the brain responsible for processing emotion. The central brain is also active, suggesting learning and memory processing.


1.3 Do Dreams Have a Function?

 Some studies have shown that sleep and dreaming are linked to learning and to repair of the body and mind.1,38  If we deprive ourselves of dream or REM sleep, the body recovers by increasing dream sleep first, recovering the loss of REM almost precisely. This suggests that dreaming is important.  It has been observed that sleep and dream deprivation causes effects such as: waking dreams (visual and auditory hallucinations); interference with memory and learning; a loosening of associations; impaired waking ability to do tasks requiring focused attention; or difficulty maintaining a straight line of thought, creating irritability and suspiciousness.  It has therefore been suggested that REM or dream sleep is associated with restoring mental well-being.

 Dream sleep seems to play a key role in revitalizing new and old experiences so that they become more permanently etched in long-term memory. Alan Hobson, at the Massachusetts Mental Health Center, shows that dreaming rehearses memory patterns, either to create new memory connections or to strengthen older fading connections (based on brain wave activity in the hippocampus).38  For example, one study demonstrated that the exact neuronal firing patterns present when rats explored a maze, were repeated precisely when the rats were in the dream sleep.38 In Israel, researchers at the Weizmann Institute found that consistently interrupting dream sleep in a night completely blocked learning, whereas similarly interrupting non-dream sleep did not.38  

 Different types of memory tasks seem to respond to different parts of the sleep cycles, with visual and perceptual tasks relating more to dream sleep.  The Weisman researchers found that “procedural memory,” or tasks requiring practice (as opposed to “declarative memory” or fact recall), was directly tied to quantity of dream sleep.  Further research by Stickgold and Walker indicated that those learning a task in the evening, who were retested after a good night’s sleep, were 15% to 20% faster and 30% to 40% more accurate than those simply tested twelve hours after learning the task at other times of the day. cited in 66

 There is some evidence of this learning effect in the dream experience itself.  On the occasions when we recall multiple dreams from the same night, it is often observed that a similar theme runs through each dream segment, but that each segment approaches that theme with a slightly different storyline, or set of characters and outcome.  Could it be that dreams act as our testing ground for practicing various scenarios on a situation, in order to find and reinforce the best solutions (strengthening the best neural paths)?

 Some research supports the hypothesis that dreams help us adapt to stressful waking events by activating habitual defense mechanisms, and by integrating the stress situation with earlier solutions to a similar problem.2  A critical step in the memory cycle is the step that matches representations of new experiences with the representations of closely related past experiences. This is observed to take place during dreaming.3  It is as if the dreams are helping us adjust to new threats and experiences, by comparing them to an internal model of how we dealt with similar situations in the past, and then making those slight reprogramming or learning adjustments that are needed to better accommodate the experience.  Thus the old saying, “better sleep on it.”  Sleep and dreaming appear to play critical roles in memory and learning, as well as in stress reduction and generally keeping our mental state in good repair.


1.4 Are Dreams Meaningful?

 In the dream state, our dreams appear similar to normal waking-life adventures. Upon waking, however, that perception changes dramatically.  We notice a lack of continuity between scenes, irrational cause and effect relationships, and a host of  bizarre or irrational combinations of dream elements or activities.

 We are conscious in our dreams, but it is a different state of consciousness from waking life. The synchronizing signals that are signs of consciousness (firing signatures of the intralaminar nuclei from various neural paths) are present, but unlike the waking state, these nuclei are not firing in response to outside stimulus.38  Instead, the brain is activating itself purely from within. But to what purpose?  Some researchers, such as Hobson and McCartney, cited in 67 contend that dreams result from our higher brain centers attempting to make sense of the activity in the lower centers that generate the dreams.  From this viewpoint, dreams could be simply be an attempt to make a rational story out of random neural activity.

 Although not totally certain at this point, some very important processing may be taking place in the dream state. As discussed above, some studies show that dream sleep plays an important role in memory and learning.  The brain centers that are active in dream sleep are also active when we are awake, but are processing experiences coming in from the external world. These centers process our perception of the external world in relationship to our internal model of that world.  They orient our physical and social perception of self to that external world, so that we can deal with it.  They stimulate our emotions in order to draw our attention to what is important and to prepare us for action in the face of a threat.  They contribute meaning and inflection to our speech.  Could it be that these same brain centers also perform a similar function in the dream state?


1.5 Dream Recall

 Earlier it was reported that some studies indicated when subjects were awakened during REM sleep, they reported dreams or dream-like experiences an average of 82% of the time.39  Assuming that most humans go through multiple REM (or vivid dream cycles) each night, it is therefore quite interesting that most of us spontaneously recall very few, if any, dreams in the morning. Hobson39 indicates that working memory is off-line in the dream state, and that the mechanisms for storing memories are diminished to non-existent.  Hobson also reports that dream recall rapidly falls off, the longer one takes to wake up after the REM period. Dreams do not occur as we awaken, as some early speculation suggested.  Dreams occur roughly over the time period we recall them having occurred. The measured length of the REM period has been correlated with both the dream report word count and the estimated length of the dream by the dreamer.39   

 Recall might relate to one’s sleeping habits.  Webb and Agnew51 found that people, who sleep longer than 8.5 hours, had 50% more REM sleep than people who sleep less than 6.5 hours. In the morning, sleep tends to be less deep and closer to waking, with longer dream periods as illustrated in figure 1.  Based on this information, you would expect more recall and longer dream reports as people sleep longer.  There are still mixed results on this, however. Whereas Backeland and Hartmann52 found it to be the case, Blagrove et. al.50 did not.  The total opportunity for dream recall may be greater as you sleep longer, but spontaneous dream recall still depends on other factors, including the dreamer’s interest in recalling dreams and attitude toward dreams (Hartmann76.)  If a person takes an active interest in recalling a dream on a particular night, or during a period of nights, they are more likely to do so.

 In a literature review cited by Lynne Hoss,45 a relationship was shown between dream recall frequency and artistic and imagery abilities. She reports that Schechter, Schmeidler and Staal (1965) tested both dream recall and creative tendencies in 100 students of art, science and engineering. There was a significantly higher proportion of dream recall among art students (assessed as more creative, and therefore using more right hemisphere process as described in chapter 3).  Recall was lowest among engineering students (attributed to the more linear, temporal thought process of the left hemisphere).  No differences between the sexes were found.  Her work also cites research by Cory, Ormiston, Simmel and Dainoff (1975) that found recall to be greater in those with greater memory capability for visual images.45  

 Spanos, Stam, Radtke & Nightingale, (1980) found that in females, dream recall was greater for those who had more ability to become absorbed in imagery and measures of creativity.  Cohen48 found a highly significant difference between dream recallers and non-recallers in ability to form clear, vivid images.  Blagrove50 found that dream recall is related to the personality factor of  “openness to experience.”  Hartmann76 indicates that there are only a few personality factors that closely correlate with dream recall including: tolerance of ambiguity, openness to experience, absorption, creativity, fantasy-proneness and ability to be hypnotized.  He states that these factors also closely relate to “thin boundaries.”

Go to Section 2

Return to The Science of Dreaming Index


Next Section

Dream Masters