Electric Dreams

New Trends in Dream Brain Research

Richard Catlett Wilkerson

(Electric Dreams)  (Article Index)  (Search for Topic)  (View Article Options)

  Wilkerson, Richard Catlett (2003 Mar). New Trends in Dream Brain Research. Electric Dreams  10(3).


The following article is really a collection of notes, re-written in a kind of summary format so that I can get a sense of the major changes occurring in the field of dream science related to the dreaming brain. I have decided to put these out on Electric Dreams as I get so many questions about the brain in sleep and this is, after all, the 50th anniversary of the discovery of REM. As the year goes on, I will organize these notes a little more coherently. If you have forgotten your sleep stages science and what happens, or are not familiar with this information, there are two appendices with summaries and detailed summaries. I haven't gone deeply into the neurochemistry of REM in the brain stem in this article and am focusing more on the general shifts in neuromodulation. However, many new changes in the neuro-circuitry of REM have occurred in the last few years and if you have in mind the older models given by Hobson, note that these are basically intact, but highly modified and expanded.

In general, this is a summary of the two articles that came out in the 2000 Behavioral and Brain Sciences journal #23. Dreaming and the Brain: Towards a Cognitive Neuroscience of Conscious States. One article by Allan Hobson, the other by Mark Solms. The drafts for these articles are available online.


Does and understanding of the mechanisms of the brain really make any difference to a dreamworker? After all, what difference does it make to me if the dream comes from the amygdala or a pre-frontal lobe? Probably none. But I would like to make the case, [without developing it very far in this article] that dreams are events made up of multiple forces. Understanding where these forces come from and where we can go with them has always been part of the dreamwork tradition.

In this article, I will look at a model in dream brain research called A.I.M. and interweave this model with radically new discoveries from brain imagining and brain lesion studies. The two major points I would like readers to get are:
1. The contributions of the higher and lower brain in dreaming.
2. There is more to the dreaming brain than just activation of areas, there is also information gating and neurmodulatory control.

*** Three major areas that change in the dreaming brain ***

The human brain (and mammal brain) can be seen as changing in three general ways as we move from wake to sleep to dream.

First, the general (A)ctivation of brain is at its highest during waking and lowest during non-dream sleep. In dream sleep, the brain is almost as active as when awake, but not quite the same way, with a shift from brain centers associated with linear thinking and calculating to areas that are connected with feeling and imagining.
Second, the gates for sensing the outer world (5 senses) and the gates that allow messages for the body to move (motor movements) from the brain to the body are the most open during wake and the most closed during dream sleep. In fact, during normal dream sleep, the only easily observable movement are (R)rapid (E)ye (M)ovements (REM) that can be seen behind a dreamer's closed eyelids. Almost all other messages from the brain to the body to move are stopped before they leave the brain, which of course protects us from moving around too much during dream sleep. In (N)on-dream sleep (NREM), we adopt sleep-postures, we can make slight shifts in our body position, and our ability to block out unnecessary noise and disturbances is less than when we are in dream-sleep. But since our brain is less activated in NREM sleep than REM dreaming sleep, the noises and lights don't disturb and wake us very easily. (note this difference again, in REM dream sleep we are less aware of the outside due to the input/output gates being shut down, but in deep sleep we are not aware due to our brain being less activated.)
Finally, the third change, the neuro-chemistry of the brain at the level of our nervous system changes from wake to sleep to dream-sleep. These neurochemical states involve the way the brain communicates with itself and our nervous system. Since they modulate various brain behaviors, they are often discussed as neuromodulators. As mentioned, these impact the overall way the brain functions, but the mind in sleep as well, just as when we take various medicines or drugs that may alter our consciousness.

*** Picking a model for viewing the dreaming brain ***

The modern science of REM based dreaming is just about fifty years old, and is already quite complex and full of controversy. After all, dream science includes the study of consciousness and unconsciousness, brain and body, sleep and wake, fantasy and reality. To grasp this complexity, scientists propose models that generalize how dreaming works. These models are then tested and revised as new data and research emerges. In this report on the dreaming brain, we will be looking at the Activation/Synthesis model developed by the Hobson group and how it has been revised in early 21st Century to include new brain studies and research made possible by brain imaging techniques, new brain function studies and new brain chemistry. The Activation/Synthesis model looks at how the lower, subcortical brain activates the higher cortical brain in REM sleep which allow the cortical brain to synthesize dreams.

I will also be toning down the causal and isomorphic parts of the Activation/Synthesis hypothesis which have caused so much controversy and are as yet highly speculative. That is, I will not be emphasizing the several hypothesis that try to such things as flying in dreams being the result of intense bursts of brainstem neurons, or paralyzed feet in the dream being the direct result of de-activated cortical areas. Other research, such as how aphasia or damage to higher visual centers that ruins a person's ability to recognize faces, will be included. Also, I will be emphasizing the Synthesis over the Activation part of the theory. Finally, I will not be giving the psi dream factor the credit it is due for the sake of brevity. Dream psi research explores alternative ways to the 5 senses that we may be in contact with others and the outer world. When I say the brain is cut-off from the outer world, I mean that we are cut off from our five senses.

*** REM Sleep Summary ***

If you are not familiar with REM and Sleep Stages, see that section below. As a quick reminder, sleep stages range from light to deep sleep. As we go to sleep, we slowly sink down into deeper stages of sleep (meaning here that the brain less activated), then periodically come up via REM (Rapid Eye Movement) dream sleep (brain more activated but cut off from outer world), then descend again. Over the course of a usual eight hour night, we will rise into REM dream sleep about 6 times, each period averaging 20 minutes of REM dream sleep, though more accurately we have longer REM periods towards the end of the night, sometimes lasting over an hour. Dreaming can occur in both REM and NREM(Non-REM sleep, stages I-IV) though traditionally we talk about REM dreaming as being longer, more vivid, and more story-like, while NREM dreams are traditionally described as being more thought-like and shorter. There is constant controversy over just how much difference there really is between REM and NREM dreams. Reports vary from 5% to 30% of the NREM dreams being indistinguishable from REM dreams. This issue will become important again as we look at the work Mark Solms and his view that REM is only one of the keys to turning on dreams. For now, I will refer to REM dreaming as a state separate from dreaming in general.

*** The Activation/Synthesis Hypothesis ***

The Activation/Synthesis Hypothesis is a fairly easy way to understand the dynamics of the dreaming brain, though it misses the richness and depth of the dream experience itself. Hobson's group proposed that during REM sleep the lower brain provides enough (A)ctivation for the upper brain to (S)ynthesize information into a dream. Further, there are 3 independent ways the brain changes that contribute to its unique states of waking, sleeping and dreaming. They are (A)ctivation of the brain sites, (I)nformation or input/output gating and (M)odulation of neurotransmitter systems.

A.I.M. Model of 3 Areas of change in sleep and dreams
[chart only works in Courier New or even spaced fonts]

------------ (A)ctivation ------ (I)nput -------- (M)odulating
Level Output Neurochemicals
Waking ------- High Open Aminergic
Sleep ------- Low dampened Aminergic-Cholinergic
REM ---------- High Closed Choinergic

*** Activation in the Upper Brain vs Lower Brain ***

We aren't conscious in sleep when our brain is not activated. Researchers used to believe that when we weren't getting enough sensory stimulation in waking life we would fall asleep. But then the reticular activating system (RAS) was found and we now know that brain is kept awake not by direct input from sensory pathways, but by tonic (longer lasting activation modulated by neurochemicals) activity in pathways from the reticular formation. This means that sleep comes from the reduction in activity from the reticular formation and wakefulness by the return of activity in the reticular formation. This system seems to be regulated by an internal clock in the hypothalamus.
Humans and other mammals are tied to the outer daily or circadian clock, the sun and to this internal circadian clock located in the hypothalamus. The suprachiasmatic nucleus of the anterior hypothalamus is the best candidate as any damage to this area change the sleep cycle dramatically and repair causes the return of normal cycles. This circadian pacemaker is also sensitive to light-dark cycles of the day but can be set or re-set to different rhythms with some discomfort, as those who get the night-shift or experience jet-lag know.
Changes to the reticular activating system that runs up though our brainstem causes changes in activation of higher brain functions. Damage or dampening of activation to a variety of particular brain areas will cause dampening of conscious activity.
In addition to the hypothalamus and the activation levels of the reticular system, another regular system engages during sleep, the REM-NREM cycle. Sleep is not single process, but rather has these two distinct phases that alternate cyclically in a very organized way through the night.
The lower brain & forebrain seems to play a critical role in the activation of REM-NREM cycle, while the forebrain and higher brain centers play a role in the formation of dreaming. REM sleep is generated by a region in the brainstem, called the pons, and adjacent portions of the midbrain. More will be said of these later.

The early presentations of the Activation/Synthesis hypothesis ran into great resistance as the Hobson group focused mostly on the Activation side of the equation. That is, they focused on the lower brain stem mechanisms that were involved in the REM state. This seemed a reasonable approach. Since the activation of the upper brain by the lower brain stem seemed to happen after cyclical phases of random nerve firings (PGO waves in cat studies), the theory was often characterized as dreams being the results of a sleepy (upper) brain doing its best it could to handle random signals from the lower brain. Allan Hobson admits that the many years of focus on the Activation side of the research led to, what he feels, this misperception of the Activation/Synthesis hypothesis. Now, new brain research on the involvement of upper brain structures in dreaming have helped to fill in the Synthesis side of the equation and allow for theories that emphasize the upper brain as more autonomous in synthesizing its own information in dream formation.

*** A.I.M. Activation, Information input/output and Modulation of Neurochemicals. ***

With the general two-part notion of lower brain activation and upper brain synthesis in dream creation, we can now look at the three major areas that change between waking, sleeping and REM dreaming through Hobson's A.I.M. model. This model tracks three general areas of brain, its 'A'ctivation levels, the 'I'nformation input/ouput gates and the neurochemical 'M'odulations that change over these states of waking, sleeping and REM dreaming.

Generally speaking, when we go to sleep the brain becomes deactivated, desensitized to outer sounds and sensations and switches over from an aminergic neurochemical system that keeps us alert and focused on the outer world to a cholinergic system that allows for relaxation. We are sleeping. Then something strange occurs, the aminergic system stops almost completely and the cholinergic system becomes hyperactive.

(To see the brain parts impacted by sleep and dreams, see):

During this time, many parts of the brain become active, the body becomes rigid, and we begin to dream (or more accurately, dreamers that are awakened from this state are more likely to report dreams and longer, richer dreams, than most other dream states.) It is as though the brain were like a computer that has been taken offline but kept running. While dreaming, it is functioning much in the same way as waking, but the inputs and outputs and connections to normal feedback from the environment are missing or dampened.

(A) Activation.

Activation includes the electrical output of the brain's surface as measured by EEG Electroencephalograms and micro-flows of blood into active areas of the brain as measured by imaging machines such as PET and MRI. This allows us to determine what areas of the brain are in operation and active. Unfortunately, EEGs only show general surface areas and only a handful of brain imaging studies have been done on dream sleep, and all of these (as of 2003) within REM. (As mentioned above, dreaming can occur outside of REM sleep and we are waiting for brain imaging studies with NREM focus as well as in dreaming, lucid dream focus.).

See "Sleep Stages: A More Detailed Summary" in the appendix for descriptions of EEG in sleep and dreams and details on what brain parts are active.

In general, when awake, our brain shows low-voltage(how high) fast pattern, which print out like the line of an eyebrow. During REM, the brain will show waves similar to waking; low-voltage, fast pattern. Specific brain areas that have been shown to be active from brain imaging studies are discussed below in the Specific Forebrain Structures Activated in Dreaming, however a general description might be as follows: The main areas activated in the upper brain during dreaming are 1. the hunting, seeking, desiring system, the 2. heteromodal 3-D imaging system and the 3. higher visual cortex. There is some evidence that these areas are activated without the regular arousal of REM, but it is clear that these areas are always activated by the lower brain in REM. Thus we might say that that REM is the main key to the driving our dream car, though there are other ways to start the car.

(I) Information input/output

Sensory input and motor output are dampened during REM, open or high I/O during waking and slightly dampened during NREM sleep. This means signals from the brain to the body are pretty much cut off and we are paralyzed during REM (some theorize so that we don't act out our dreams) with some exceptions, such as eye movement, flow of blood to the genital regions increases, and a few other minor movements.

"I" is measured by EMG postural muscle tone (how relaxed our body is) and EOG, eye-movement activity.

Sensory isolation during REM comes from the inhibition of the Ia afferent terminals (endings of the sensory nerves that form synapse with neurons in the brain itself). The source seems to be in the brain stem, the pontomedullary reticular formation that hyperpolarizes the motoneurons (makes them less responsive to commands from the brain to act, ie, motor commands). Loss of muscle control is from tonic postsynaptic inhibition of spinal anterior horn cells by the pontomedullary reticular formation.

In general, most of the outgoing motor messages from the brain are cut off to the body at the medulla and incoming sensory data from the five senses are inhibited. This is not a black and white situation. Alan Worsely, for example, reports that during the first lucid dream signaling experiments, he was able to vibrate his hands from dream lucid dream sleep. This indicates that the outgoing motor-muscle messages are dampened rather than being fully shut off.
Of course, the eyes move rapidly during REM dreaming and many structures and neural routes have been suggested between the lower brain and eye movement.

In REM Behavior Disorder (RBD) people act out there dreams. This is not sleepwalking, which occurs in NREM. "The inhibition of movement or motor output, which normally quells the movement commands of dreams, is only quantitatively greater than the excitation of neurons that is the embodiment of these commands. If either inhibition declines or excitation increases, or both, movement will result." (p96 Dreaming Brain, Hobson)

In other words, when there is an imbalance in the brain stem (due to neurological problems or perhaps in lucid dreaming to active pre-frontal lobe commands) one can break the REM-barrier! Hobson reports in the Dream Drugstore (2001) one patients flailing arms, hitting his bed partner, only to wake up and recall having to turn the wheel on his car to avoid a cliff. Another patient dreams of swimming and crawls right off the bed.
The suspected cause is an imbalance in dopamine, a neurotransmitter involved in one brain system with the condition of parkinsonism, and in another part of the brain with the activation of the hunting, seeking, desiring system. (Solms 2000) Hobson reports that sometimes the prolonged use of anti-depressants corresponds to the RBD condition.

As mentioned above, sleepwalking occurs in NREM, as well as sleep-talking and tooth-grinding. People awakened from these activities don't recall dreaming. They are automatic behaviors coming from the lower brain areas called motor pattern generators. Hobson says its fine to wake these people up without psychological damage, if you can. They are usually in stage IV deep sleep and very difficult to arouse.

(M) Modulation

This is the strength of chemical systems modulating the brain. For Hobson, this is measured in the ratio of cholinergic to aminergic neruomodulator release. In the Reciprocal Interaction Model of REM, these two systems switch in waking and dreaming, with aminergic systems dominant in waking and the cholinergic system dominant in dreaming. In NREM, all three tend to be de-activated.

More accurately, in waking, the aminergic system is at its height of influence and inhibits the cholinergic system. As we go to sleep, the aminergic inhibition loosens it control slowly and the cholinergic system slowly gains in strength. In REM the aminergic inhibition is shut off and the cholinergic system is at its peak of influence.

Other researchers, like Mark Solms, feel that neuromodulation of dopamine to be more important to upper brain structures involved in dreaming (the synthesis part of the model), while cholinergic systems have more to do with only one of many activation systems.

Very little research with humans have been done in this area, but Hobson feels quite confident that this physiology that is common to all other mammals will be also be at work in humans.

*** Specific Forebrain Structures Activated in Dreaming ***


To summarize before looking into the specific brain areas involved in dreaming, the dreaming brain appears to have its own (M) neuromodulatory system that involves [at the level of the brainstem/Pons] a shutting down of the aminergic system and activation of the cholinergic system. The thalamus (basal forebrain) and amygdala are cholinergically modulated. The cortex is aminergically demodulated, especially in terms of dampening recent memory and orientation.

Activated Upper Brain Areas in Dreaming

The dreaming brain shows (A)ctivation of many areas as in waking, with the major exceptions of the prefrontal cortex (linear thinking, calculating) and the primary (V1 and some of V2) visual centers, though higher visual centers are activated. This makes sense as V1 is where information from an eye would first go if one were awake. At the level of the brainstem, the pontine tegmentum is active, involving reticular information (general arousal) the PGO system (may initiate REM) and activation of cholinergic centers (sleep and dream neuromodulators). There is particularly high activation of the amygdala and paralimbic cortex (Emotion and Recent Memory). The parietal operculum (visual-spacial imagery) is activated.

Input-Output Gating

(I)nput-output gating is in effect in the dreaming brain. At the level of the lower brain stem motor output is blocked, leaving the body paralyzed. Sensory input is blocked, making the outer world unavailable through the five senses. Hobson theorizes from cat studies and newer evidence that the PGO system is turned-on, producing input of fictive visual and motor data from the geniculate bodies to the occipital cortex. That is, as the parts of the brain that deal with motor movements and sense data are turned on, we begin to be able to dream about movement and sensorial scenes.

Upper Brain Activation and Synthesis

This may be a good place to give a summary of the areas in the dreaming brain that Mark Solms research has revealed. Solms feels that the upper brain can synthesize dreams without the help of the lower brain stem REM system.


The paradigmatic assumption that REM sleep is the physiological equivalent of dreaming is in need of fundamental revision. A mounting body of evidence suggests that dreaming and REM sleep are dissociable states, and that dreaming is controlled by forebrain mechanisms.

Solms combined recent neuropsychological, radiological and pharmacological findings with his own brain damaged patients and other extensive neurological research in the past to suggest that the cholinergic brainstem mechanisms which Hobson's group shows control the REM state can only create dreams with the help of a second, probably dopaminergic, forebrain mechanism that activates a series of higher brain systems. Hence, Solms proposes a Dream-on instead of the Hobson group's REM-on theory of dreaming. In the Dream-on theory, dreaming can be initiated by many influences outside of REM activation.

In Solms theory, dreaming begins in the higher brain when a particular area of the forebrain is activated, the mediobasal frontal cortex. Here the hunting, seeking, desiring, wanting system is deeply networked with the limbic system (emotions, sensory info) and mesocortical dopamine systems. There are deep connections of dopanminergic cells from this ventral tegmental area to the hypothalamus, the septal area, the cingulated gyrus and the frontal cortex, and amygdala. In other words, this frontal cortex area of motivation connects with many other parts of the higher brain, the sensory brain and the emotional brain.

When activated in sleep (by REM, drugs, seizures and perhaps other systems) the extensive connections begin a sequence of activation that includes the (I) input/output gating of the motor cortices (M) a dopamine modulation of brain in general and (A) and activation of the emotional systems, the limbic system (sensory, emotions), the PTO junction or inferior parietal cortex (heteromodal imagination and 3-D space), and the medial-occipital temporal cortex (higher visual centers).

Interestingly, the higher visual centers can be destroyed and we can still dream, through with noticeable differences (such as missing faces in aphasia). But other areas seem essential to dreaming. Lesions in the PTO junction where we create or have heteromodal, 3-D space sense is essential to dreaming and no dreams are reported from patients with lesions in this area, even after many years follow up. Also, extreme damage to the above mentioned ventral mesial quadrant of the frontal lobes removes any dreaming (or reports of dreaming) from the patients. Solms theorizes that just like the patients of old who have had leucotomies (lobotomy of this area), they just can't reach the arousal level needed for dreaming. Patients can still perform acts in waking with lesions in this area, but only upon request, as they lose all initiative to act on their own volition. I asked Solms if this couldn't just be lack of motivation to remember, and he didn't feel it was a memory issue as all the memory systems are intact and the patient's memories function perfectly well in other situation. Still, I wonder how many dreams I would recall in the morning if I lacked the motivation to do so.

Recent brain imaging supports this theory that dreaming involves very specific brain structures. These *activated* structures include anterior and lateral hypothalamic areas, amygdaloid complex, septal-ventral striatal areas, as well as the infralimbic, prelimbic, orbitofrontal, anterior cingulate, entorhinal, insular and occipitotemporal cortical areas. *Deactivated* structures include the primary visual cortex (where waking eye information would go, not the same as the activated higher visual centers) and dorsolateral prefrontal cortex (the calculating part of the brain).

Hobson has accepted much of Solms research, particularly on the specific higher brain areas that are activated during dreaming sleep. But Hobson doesn't feel that REM activation can ever be separated from other aspects of dreaming and is still holding out on whether or not upper brain functioning during dreaming is modulated by dopaminergic systems. (A separate dopamine system from the one often related to Parkinsons).

Solms feel a variety of research lines are converging on this same issue of the dopaninergic system in the forebrain. see:

*** Summaries of specific brain areas activated and deactivated during dream sleep. ***]

Please use figure 7 from the online Hobson article to locate the following brain structures. This section is unlikely to make any sense without the picture.


++++ Zones 1 & 2, figure 7. (Subcortical) Ascending arousal systems : 1. Pons and midbrain RAS and nuclei. PGO source. Arouses and activates brain, allowing for consciousness and eye movements. 2. Diencephalic structures (hypothalamus, basal forebrain). Autonomic and instinctual function, consciousness modulation.

++++ Zone 6, figure 7. (Subcortical) Thalamaocortical relay centers and thalamic subcortical circuitry. Thalamic nuclei (e.g. lateral geniculate body). Relays sensory and pseudosensory information to cortex.
In NREM sleep, corticothalamic waves suppress perception and mentation, but this process is reversed in REM. In REM, the thalamic nuclei activate sensorimotor parts of the brain and fill these parts of the brain with general activation. Hobson feels this may present basic elements of dream scenes in the form of pseudosensory information.

++++ Zone 3, figure 7.(Cortical and subcortical) Limbic and paralimbic structures.
Anterior limbic structures (amygdala, anterior cingulate, parahippocampal cortex, medial frontal areas). Emotional aspects of dreaming, emotional coding, goal directed behavior, movement,. For example, the amygdala when activated is correlated with anxiety and high emotions, and the amygdala activates the anterior cingulate, right parietal operculum. Deactivated are the prefrontal cortex, parietal cortex and precuneus.
the anterior cingulate if related to emotional features in waking and dreaming in integrating emotion with fictive actions.

As mentioned before in the section on Mark Solms research, this area also includes motivational centers without which we would not have access to the hunting, seeking, desiring, wanting part of ourselves.

Hobson feels this points to the notion that emotions are more the shaper of dream plots than reaction to events in dreams being the primary force driving emotions as in waking life.

++++ Zone 5 in figure 7. (Subcortical) Basil Ganglia. Motor initiation and control centers. Hobson feels this lower brain area is responsible for the modulation of movement in dreams and even adds specific features as vestibular sensations. That is, the sensation of fictive dream movement in our dreams.

++++ Zone 11 in figure 7. (Neocortical) visual association cortex. Higher visual processing centers that contribute visual information to dreams. We can dream even when this area is damaged, but our dreams will be impacted, as in the loss of face recognition in aphasia when the fusiform gyrus is damaged. At the same time, the primary visual centers (V1 and part of V2) are deactivated. This makes sense as the eyes are closed.

++++ Zone 9 in figure 7. (Neocortical). Inferior parietal lobe. Brodmann's Area 40. Spatial integration of heteromodal input. Solms refers to this area as the PTO junction (Parietal-Temporal-Occipital) and has shown that it is essential for dreaming, allowing us to imagine inner space and without it, all dreaming ceases. Also, it coordinates heteromodal information of all types. As Hobson writes, it "may generate the perception of a fictive dream space necessary for the global experience of dreaming."
Of interest to left-brain/right-brain theorists, PET studies of this area during REM show that much of the parietal lobe is deactivated, and just this right parietal operculum activated. That is, in some studies, the right is more important than the left in this area during dreaming.

++++ Zone 4 in figure 7. (Neocortical- deactivated) Dorsolateral prefrontal cortex, or the executive association cortex. Prominent deactivation in the frontal cortex. This is the executive or reasoning part of the brain and the part that we use to do math, think linearly and calculate. Hobson feels this may contribute to many of the "dream deficiencies" such as memory loss, shifts in scenes, disorientation. It will be interesting to see if this area of the brain is more activated in lucid dreaming or not. Having this part of the brain offline may contribute to better facilitation of emotional and memory consolidation processes.


*** Final Summary - How the brain works during dreaming ***

In terms of the process of dreaming at the level of the brain/body, we have learned quite a bit since the discovery of REM 50 years ago by Aserinsky and Kleitmann in the Chicago University sleep labs. REM or Rapid Eye Movement sleep occurs on a regular cycle about 20 minutes every 90 minutes of sleep. (More accurately, we have shorter REM the first part of the night, and longer REM periods, up to two hours, towards the end of the night). People often report dreams if awakened from REM.
Now we look at three different levels brain dreaming, the activation of various sites in the brain, the gating or input/output during Wake/Sleep/ REM stages and the different neurotransmitters that are impacting these stages.
In brief, when the sleeping person enters REM sleep, much of the mind that was quiet "wakes up", the dominate neurotransmitter changes from aminergic to cholinergic washes, and the output from the brain is cut off at the level of the lower brain stem. That is, messages from the activated brain go out to the body as in waking, but never make it there and so we get a kind of REM paralysis. Two areas of the brain that don't wake up are the parts of the pre-frontal cortex that one usually uses to calculate the lunch bill, and the primary visual centers used during waking site. (Higher visual centers are still activated. Its unclear still what brain parts are activated during lucid dreaming).

According to Hobson, this whole cycle is started by the changes occurring regularly in sleep in the brain stem. Mark Solms sees this brain stem activation as only one of the ways the brain starts its dreaming cycle. Solms focuses on the higher brain in dreaming and sees the beginning occurring in the frontal part of the brain that is our hunting, seeking, goal oriented center. Without it, (in damaged brain patients) there just isn't the motivation to dream or recall dreaming. From there, the activation crosses over to the very important PTO junction between our Occipital, Parietal, Temporal lobes, a place that might be described as necessary for a human to have any kind of spacio-temporal imagination. Without it, (in damaged brain victims) there is no dreaming reported. Finally the activation occurs in the higher visual centers. We can dream even without activation of these visual centers, but its unclear just what kinds of dreams one can really have. Patients with partial damage, causing for example aphasia, can't recall faces in people in their dreams.

Spontaneous or General?

Hobson sees this whole process modulated by the lower brain stem and cholinergic neurotransmitters. Solms sees the brain stem as peripheral to dreaming, as epilepsy and other events in NREM or Non-REM sleep can stimulate dreaming as well. Solms hypothesizes that the main neurotransmitter is serotonin. Either way, it is REM sleep from brain stem which seems to operate as a regular starting mechanism for the activation of the higher brain, though other spontaneous dreaming may occur outside of REM.



APPENDIX 1 *** REM and Sleep Stages ***

We live on a planet that is light half the day and dark the other half. Creatures adapt to this two-part cycle, active and competitive in one, resting and asleep in the other. But the night is not as inactive as one might predict for some of its sleeping creatures. Almost all mammals experience in sleep complex changes in brain activation levels, sensory input, motor output and brain chemistry. In humans these changes often set the brain-body conditions in which we experience dreams.

Beginning of Contemporary Dream Science in REM (Rapid Eye Movement Sleep)

In 1953 at the University of Chicago, Nathaniel Kleitman and his student Eugene Aserinsky connected eye activity in sleep to dreams. (Aserinsky & Kleitman, 1953) Dr. Kleitman had been studying sleep difficulties in infants and wanted to explore the slow rolling eye movements that babies have at sleep onset. He had his student Eugene Aserinsky watch these movements of sleepy infants. What surprised Aserinsky and changed the notion of sleep forever, was the occasional occurrence of very rapid movements of the eyes at various times during the sleep cycle. Though the eyes remained closed, they moved just as if the child was awake and outside playing games. Aserinsky and Kleitman then monitored adults and found the same thing, and that these eye movements lasted anywhere from three to fifty-five minutes (Van De Castle, 1994).

Since the movements appeared as if the sleepers were scanning a scene, they decided to awaken them and ask what they were looking at. They were, more often than not, dreaming. When they woke sleepers up when there was no eye movement, they rarely reported dreams. These discoveries were reported in _Science_ on September 4th, 1953 and again in an expanded article in 1955. It was the beginning of what is now 40 years of contemporary dream research in the sciences.

While Aserinsky finished his medical program and left the labs, William C. Dement (1976) filled his place and soon was able to characterize sleep in stages. The REM state is different physiologically than waking or other kinds of sleep. During REM sleep, there are irregular patterns in breathing, heart rate and blood pressure. Our muscles are tense, though they can twitch and jerk. Most of the motor commands from the brain to the muscles are cut off during REM above the neck.

Stages of Sleep

Although sleep stages are different in every individual and vary from night to night and differ widely from childhood to late adulthood, some generalizations have been observed.

After a few minutes of drifting we slide into deeper and deeper levels of what is called NREM or Non-REM sleep. The brain’s waves get wider and slower. After an hour or two the first REM period begins and lasts a few minutes. Then we sink back into deeper and deeper sleep. This cycle occurs about every 90 minutes. Towards the end of the night or sleep period, the REM periods get longer and we don’t sink into quite as deep of sleep.

Traditionally, three kinds of measurement used to determine the stage and level of sleep are:

1. EEG: The electroencephalogram to determine electrical activity on the surface of the brain. Short dense fast desynchronized waves during waking and dreaming, tall, wide synchronized waves during NREM sleep.
2. EOG: The electrooculogram. To measure eye movements which produce REM.
3. EMG: The electromyogram. To measure muscle tone.

Now other measurements include the brain chemistry, EKG or heart rate, respiration and PHG or genital arousal. Newer recording equipment such as the MRI, PET and other digital imaging equipment are slowly being used in dream research. These techniques take advantage of the fact that when a particular area of the brain is active, there are micro-fluctuations of blood flow in that area.

Sleep Stage Summary:

Comparing REM with waking we find many similarities. About an hour or two into sleep, people move back up through states three, two and one, and often enter the first REM stage of the night. REM sleep is sometimes called "paradoxical sleep" because it has characteristics of both light and deep sleep. The first REM period of the night usually lasts only a few minutes. Then people sink into the deeper stages of sleep again. As the night progresses, more REM periods occur and become longer and longer. Near the end of a sleep period, they can last for an hour or more. The NREM or non-REM sleep times are shortened as the night goes on.

By the end of the night, we usually have stopped having state 4 sleep. Near the end of the night (or sleep period) we rotate between stage two at bottom and up to REM.
It is easy to get dream reports from people awakened from REM, but people can dream in any stage. Sawtooth waves occur in the EEG (electroencephalogram, a surface brain activity measurement) and eyes move rapidly back and forth. Messages from brain are cut off at the brain stem and never reach the body. The body’s heating system is regulated more like a reptile and cannot heat or cool itself. It assumes the temperature of the surrounding room. Part of understanding that the REM state is different is that it is a *physiologically* different state than waking or other kinds of sleep. During REM sleep, there are irregular patterns in breathing, heart rate and blood pressure.

APPENDIX 2 *** Sleep Stages: A More Detailed Summary ***

NREM Sleep Stage 1.

Wake-Sleep Transition: As we lie down and close our eyes, (if we are tired) we begin to de-activate and move into low voltage, mixed-frequency EEG brain activity.
People awakened from this sleep stage often report just barely being asleep, or just about to fall asleep.• Short dreams, or dreamlets may be reported. Body jerks and wandering thoughts can occur.• This sleep stage usually lasts 3- 12 minutes.

• EEG: tight, fast Beta waves are replaced by looser, slower Alpha waves characteristic of a meditating mind. Soon Theta waves [4-8 Hz, (Frequency or how fast) 50-100 ΅V (Amplitude or how tall the waves peak)] begin to appear.

(I)information input/output:
• Reactions to outside stimulus diminish. We stop noticing a lot of the noise and lights.•

(M)odulation of neurochemical systems
• The daytime aminergic system begins to wane and slowly stops inhibiting the cholinergic system which slowly starts coming online.

NREM Sleep Stage 2.

• Not to hard to wake people here, but they usually report being really asleep.• Lasts 10-20 minutes

(A) • EEG: Sleep spindles appear. That is, twice as slow Theta waves. Occasional spikes called K-complexes and the beginning of large slow delta waves.
[4-15 Hz , 50-150 ΅V ]

(I) • EMG: Muscles have tone or tension. • Reactions to outside stimulus diminish. Unlikely to notice noise and lights unless unexpectedly strong

(M) • Aminergic neuromodulation system continues to loose control and cholinergic system gains more control.

NREM Sleep Stage 3.

• Lasts about 10 minutes

(A) • EEG: Slow waves [ 2-4 Hz, 100-150 ΅V
] and Delta Waves. A little less than half the waves are large, slow delta waves. Spikes and K-complexes occur, but not as much. Slow waves + spindles + K complexes

(I) • EMG: Muscles have tone. • Reactions to outside stimulus unlikely unless strong or salient (mother hearing child's call or we hear our name). Unlikely to notice noise and lights unless unexpectedly strong

(M) • Aminergic neuromodulation system continues to loose control and cholinergic system gains more control.

NREM Sleep Stage 4

More slow-wave activity in the EEG readings and overall neuronal activity at it lowest. Brain temperatures lowest and sympathetic outflow, heart rate and blood pressure down. Stages 3 and 4 in humans are sometimes called slow-wave sleep.

• Sleepers hard to awaken. Children my take several minutes to awaken.• Combined with stage 3, Lasts 40-90 minutes.

(A) •EEG: Delta Sleep. More than half the waves are large, slow delta waves [0.5-2 Hz,
100-200 ΅V]

(I) .• EMG: Muscles have tone. • Sleepwalking, sleeptalking, night-terrors, bedwetting in children.

(M) • Aminergic neuromodulation system continues to loose control and cholinergic system gains more control.

REM Sleep Stage:

Just exactly what starts REM sleep is complex and partially still being investigated.
A system of neurons generating the EEG, eye movement, twitches and underlying muscle atonia of REM sleep have been identified in the brainstem . This system utilizes adrenergic (noradrenergic and serotonergic) REM sleep-off neurons, GABAergic, cholinergic, glycinergic and glutamatergic REM sleep-on cells as well as other neurons.

In REM or Rapid Eye Movement sleep, the EEG looks similar to stage 1 NREM and waking. Because it resembles waking, REM is often called "paradoxical" sleep.
In REM there are bursts of neural activity, expecially in the Pons. These bursts generate high-voltage spike potentials, the ponto-geniculo-occipital or PGO spikes. The PGO spikes are named after structures in which these spikes are most detected (the pons, lateral geniculate nucleus, and occipital cortex). PGO spikes are one of the phasic or short-lasting events of REM sleep, including eye movements and cardio-respiratory irregularity.

The overall activity of the brain increases, and so the brain temperature and metabolic rate are high, equal to or greater than during the waking state. Atonia occurs (loss of muscle tone or outgoing motor commands to muscles) though small, phasic twitches occur and the skeletal muscles controlling the movements of the eyes, middle ear ossicles, and diaphragm are not atonic. The pupils are constricted (miosis), reflecting the high ratio of parasympathetic to sympathetic output to the pupil. Genital arousal regularly occur during REM sleep. There is a reduction in homeostatic mechanisms. Respiration is relatively unresponsive to changes in blood CO2, and response to heat and cold are absent or greatly reduced. Thus the body temperature drifts toward room temperature as with reptiles.

(A) • EEG much like waking [15-50 Hz < 50 ΅V ] and stage 1
•EEG gamma frequency 30-80 cycles per second "that has been touted as denoting sufficient temporal coherence among the widespread neuronal circuits of the context to permit the binding necessary for the unification of conscious experience. "

Pontine tegmentum: activated retircular formation, PGO system and cholinergic system.
Amygdala & paralimbic cortex : activation of emotional (quantity) and remote memory.
Parietal operculum (PTO junction) : activated visuospatial imagery
Prefontal cortex deactivated: volition, insight & judgement and working memory all deactivated.

EKG: Irregular heatbeat compared to NREM

(I) • EOG: Rapid Eye Movements back and forth rapidly. Sometimes measured by strain gauges as well. EMG: muscles loose and relaxed. • Active suppression of senory input and motor output. That is, stimuli from the outer world is dampened and messages from the brain to move are cut off at the brain stem. (Eyes are an example of the few outgoing nerves not dampened, and hence REM)
Motor output blocked: real action dampened
Sensory input blocked: outer world data unavailable
PGO system turned on: fictive visual & motor data generated

• Respiration is less regular than NREM

(M) • Aminergic demodulation, Cholinergic control. (suppression of firing by locus coeruleus and raphe neurons). " REM-on cells are postulated to occur via disinhibition (resulting from the marked reduction in firing rate by aminergic neurons at REM sleep onselt) and through excitation (resulting from mutually excitatory cholinergic-noncholinergic cell interactions within the pontine tegmentum" 138

Aminergic demodulation (loss of waking mental tone) may be a more or less direct cause of the difficulty in moving dreams from short to long term memory, as the attention needed to code memory is difficult with aminergic demodulation.

Thalamus basal forebrain & amygdale cholinergically modulated.
Cortex aminergically demodulated: recent memory and orientation down.
Pons: switch from aminergic (now off) to Cholinergic neurons (now on)

Schematic summary of REM:

"EEG dysychronization results from a net tonic increase in reticular, thalamocortical, and cortical neuronal firing rates. PGO waves are the result of tonic disinhibition and phasic excitation of burst cells in the lateral pontomesencephalic tementum. Rapid eye movements are the consequence of phasic firing by reticular and vestivular cells; the latter directly excite oculomotor neurons. Muscular atonia is the consequence of tonic postsynaptic inhibition of spinal anterior horn cells by the pontomedullary reticular formation. " From Hobson et al., 2000 Behavioral and Brain Sciences 23

PGO waves:
In cat studies, oncoming REM seems to come from the lateral geniculate bodies of the thalamus, corresponding to the depolarization of the geniculate neurons by excitatory impulses arising in the pontine brain stem, and depolarization of neurons of the reticular formation and the PPT pedunculopontine region. The PPT, a cholinergically modulated area, is thought to be the origin of the process that initiates REM in the brain stem. signals originate in the pons (P) and radiate to the geniculate bodies (G) and the occipital cortex (O).

REM begins when PGO waves become cholenergically hyperexcitable, a condition that is regulated by the inner circadian clock in the thalamus.
More specifically, the " of serotonergic inhibition and neuromodulation that results from the "Don't Act Now" signals sent down into the pons from the hypothalamic circadian clock. "

This mode of active signals without input/output gating means we have a lot of "fictive movement" or movement hat is centrally commanded but peripherally inhibited.



Hobson, Allan J. (2001). The Dream Drugstore: Chemically Altered States of Consciousness. MIT Press: Cambridge, MA

Hobson, J. Allen, Pace-Schott, E. and Stickgold, R. (2000)
Dreaming and the Brain: Towards a Cognitive Neuroscience of Conscious States
Behavioral and Brain Sciences 23 (6): 793-842
Available online:

Hobson, Allan J. (1995). Sleep. New York, Scientific American Library.

Hobson, Allan J. (1988). The Dreaming Brain. New York: Basic Books, Inc.

Kryger, M., T. Roth, et al. (1994). Principles and Practice of Sleep Medicine. Philadelphia, W.B. Saunders Company.

Rechtschaffen, A. and A. Kales (1968). A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. Wasington, D.C., NIH Publication 204.

Sleep and Dreaming Rechtschaffen, A. and Siegel, J.M. Sleep and Dreaming. In: Principles of Neuroscience. Fourth Edition, Edited by E. R. Kandel, J.H. Schwartz and
T.M. Jessel, 936-947, McGraw-Hill, New York, 2000.

Solms, Mark (1997). The Neuropsychology of Dreams: A Clinico-Anatomical Study. Mahwah, NJ: Lawrence Erlbaum.

Solms, Mark (2000), Dreaming and REM sleep are controlled by different brain mechanisms, Behavioral and Brain Sciences 23 (6): 843-850.
Available online:

BBS Special Issue: Sleep and Dreaming