A Brief Review of the Psychological Literature on Dreaming Dan Lee ------------------------------------------------------------------ Abstract A great deal of physiological and psychological research has been done on dreams. The advent of tools to measure brain waves, eye movement, and muscle tone have enabled the field of dream research to make dramatic advances in the last 40 years. Dream research has also contributed to theories on how the mind works, and neural network theory has provided the first memory models to adequately account for the associative and emotional characteristics of dreams. In addition, recent research on the phenomenon of lucid dreaming is changing our understanding of consciousness itself, and of the relationship between consciousness and sleep. Yet despite the progress in understanding REM sleep and dreaming, relatively few therapeutic applications have been developed. ------------------------------------------------------------------ Table of Contents Introduction Overview of REM sleep theories REM sleep physiology Theoretical studies Studies of dreaming and memory Neural network models Field independence and dreaming Intense dreams Nightmares Archetypal dreams Background to lucid dreaming Lucid dreaming Understanding intense dreams Conclusions ------------------------------------------------------------------ ------------------------------------------------------------------ Introduction Until the 1950's, sleep was considered to be a time in which the brain was quiet and inactive. Dreams were only vaguely understood, and were thought to be either random perturbations of consciousness, symptoms of repressed primal instincts, or, among some, visions provoked by evil spirits. The perception of sleep as a mentally dormant state changed radically when researchers began using the electroencephalograph, or EEG, to monitor sleep. The EEG is a device that measures electrical activity in the brain. When the EEG was used to measure brain activity in sleeping subjects, it came as a surprise to many researchers that the brain is sometimes just as active in sleep as it is in waking. In 1953 researchers made the connection between REM sleep and dreaming - and opened the door to modern sleep and dream research. There are five stages of sleep. We pass through each of the stages several times a night in 90-100 minute cycles. Each stage is marked by a unique pattern of EEG waves. The EEG measures small changes in electrical potential through electrodes placed on the scalp. The amplitude (height) of the waves on an electro- encephalograph reflects the strength of the electrical activity in microvolts. The frequency (distance between peaks) of the waves reflects how much neural activity is occurring. EEG frequencies are measured in Hertz (Hz), or cycles per second. The lowest rhythm, approximately 1 to 3 Hz, is termed a delta rhythm. The next highest, from 4 to 8 Hz, is called a theta rhythm. Finally come the alpha (8 to 13 Hz) and betarhythms (13 to 30 Hz). Other tools used in sleep research are the electrooculogram, or EOG, which records patterns of eye movement, and the electromyogram, or EMG, which measures muscle activity. Eye movement and muscle tone also change throughout the stages of sleep, the most dramatic changes appearing in the transitions between REM and non-REM (NREM) sleep. In the normal waking state, our brain shows EEG patterns of alpha and beta waves. We can calm the mind quickly simply by relaxing comfortably with the eyes closed. Closing the eyes shuts off visual input from the outside world, and by resting quietly in this manner the brain will settle into a pattern of alpha waves (Hobson, 1989). In this relaxed waking state muscle tone is high, and eye movement irregular. As we enter stage 1 of sleep, the brain begins to show theta waves. We may also begin to experience "vivid, brief dreamlets called hypnagogic (from Greek, meaning 'leading into sleep') imagery" (LaBerge, 1990, p. 21). These images foreshadow the dreams of REM sleep, but healthy individuals typically do not enter the REM stage until the end of the first full cycle of sleep. As stage 1 progresses into deeper sleep, electrical activity in the brain slows down, and we enter stage 2. Eye movement is less pronounced, and brain activity appears to be more random. It is during stage 2 that the first true indicators of sleep appear on the EEG. These are bursts of neural activity known as sleep spindles and K complexes, which show up on the EEG as tight patterns of rapid wave movement. Sleep spindles continue to appear in stage 3 as the EEG shows a slower pattern of theta waves that gradually gives way to delta waves. Stage 4 sleep is characterized by slow, strong high-amplitude low-frequency delta waves. This is the deepest level of sleep, when the brain is at its lowest level of activity. As we come out of delta sleep, the brain quickly moves back up through the stages to stage 1. Sometimes we arouse and awaken at this point, but more often we enter into REM sleep. In contrast to the deep, quiet delta sleep, REM sleep is marked by a high level of brain activity. The low amplitude recorded by the EMG during REM sleep indicates a state of very low muscle activity during the time the brain is most active, a phenomenon known as sleep paralysis. LaBerge (1990) provides a good description of the REM state: The transition from quiet to active sleep is quite dramatic. During the active sleep phase, commonly called rapid eye movement or REM sleep, your eyes move rapidly about (under closed eyelids, of course) much as they would if you were awake. Your breathing becomes quick and irregular, your brain burns as much fuel as it does when you're awake, and you dream vividly. If you're a male, you probably will have an erection; if you're female, increased vaginal blood flow. While all this activity is happening in your brain, your body remains almost completely still (except for small twitches), because it is temporarily paralyzed during REM sleep to prevent you from acting out what you're dreaming (p. 20). LaBerge points out that REM sleep is unique in that it is both a state of consciousness and a specific physiological state. This state involves "the cooperative activity of a number of special purpose brain systems. For example, independent neural systems cause muscular paralysis, blockade of sensory input, and cortical activation. When these three systems are working together, your brain will be in the state of REM sleep, and you will probably be dreaming" (LaBerge, 1990, p. 108). This blockade of sensory input is what makes it possible for us to sleep and dream despite such noisy irritations as nearby traffic or buzzing alarm clocks. The suppression of this input is accomplished in at least two ways. First, signals from the brain stem inhibit external impulses from the peripheral nervous system from reaching the central nervous system. In his book "The Dreaming Brain," J. Allan Hobson (1988) writes: Such pre-synaptic inhibition has been recorded at sensory-relay centers throughout the brain. The physiologist Ottavio Pompeiano...has shown that the nerve terminals from primary sensory neurons feeding information either into the spinal cord, or into certain brain stem and thalamic sensory-relay nuclei, are depolarized by signals coming from the brain stem. (p. 206) Second, external sensory signals that do get through are simply tossed into the whirling cauldron of messages already being generated by the REM state, where they must compete for attention among circuits that are already overloaded. Hobson calls this effect "occlusion", and states: The net effect of occlusion is the outright neglect of external stimuli or the facile incorporation of them into the internally generated information stream of dreaming. In a REM-sleep dream, I once interpreted a buzzer (which timed the EEG records in our sleep lab) as a telephone ringing. Such stimulus incorporation into dreams not only helps preserve sleep, but also provides insight into the temporal aspects of the brain-mind interaction in sleep. (p. 206) The phenomenon known as "REM rebound" occurs when REM sleep is interrupted or inhibited for some period of time. REM sleep deprivation studies have demonstrated that the brain will attempt to recover its lost REM time. In one study, volunteers were awakened during the night as soon as they entered REM sleep. Over ten successive nights, the number of times the subjects entered REM sleep soared, from three or four times the first night, to over 30 times on the final night of REM deprived sleep. "The important implication of these results is that there is a very real and pressing need, or drive, for humans to enter REM sleep. The longer we try to prevent REM sleep, the harder our brain tries to get it" (Hobson, 1988, p. 112). Overview of REM sleep theories Clearly there is some physiological reason or need for REM sleep, but what is it? Several theories have been proposed, ranging from the purely mechanical to the deeply spiritual. One theory focuses on the development of the brain in the fetus. Newborn infants spend 50% of their time in REM sleep (Bloom & Lazerson, 1988, p. 188), and the human fetus spends up to 80% of its time in REM (Hobson, 1988, p. 195). A well known study has demonstrated that mental stimulation greatly enhances neural development (Resenzweig, Bennett, & Diamond, 1972). These observations have led to the idea that the high level of mental activity during REM sleep provides the infant brain with its own, internal mental stimulation, promoting and enhancing neural development. Some envision dreams as simply the brain's technique for memory consolidation and clean-up activity (Crick & Mitchison, 1983), others see dreams as an emotional safety valve, emphasizing the rich emotions and deep personal meaning that dreamers sometimes experience (Cartwright, 1990). Still others regard dreams as the workings of a holistic neural network in which connection strengths involving the entire brain are expressed in similar ways during both waking and sleep (Globus, 1989). One of the more interesting areas of research is the phenomenon of lucid dreams. A lucid dream is a dream in which the dreamer is aware of the fact that he or she is dreaming, and is also able to willfully control the content of the dream. Tart (1979) describes the lucid dream as follows: Lucid dreaming is an altered discrete state of consciousness characterized by the lucid dreamer experiencing himself as located in a world or environment that he intellectually knows is "unreal" (or certainly not ordinary physical reality) while simultaneously experiencing the overall quality of his consciousness as having the clarity, the lucidity of his ordinary waking discrete state of consciousness. ...[this characterization of] the overall configuration of his consciousness as practically identical to his waking state by the lucid dreamer is the crucial defining element of lucid dreaming. There does seem to be a difference between realizing that you are dreaming and actually becoming conscious within the dream. Often the realization that you are dreaming wakes you up into the real world rather than causing you to become lucid in your dream. (p. 255) Since dreaming itself is considered to be a chance event in our culture, very few people ever attempt to exert control over their dreams. The ability to attain lucidity within a dream appears to be limited to a few individuals who simply seem to have a talent for it. However, some dream researchers have been able to identify techniques that have proven useful in learning to achieve some control over dreaming (LaBerge, 1980; Tholey, 1983; Gackenbach & Bosveld, 1989). The prospect of being able to directly access the unconscious mind through lucid dreaming has excited some researchers. Lucid dreaming appears to be a technique that is not being discovered, but rather, rediscovered. Edward Bruce Bynum (1993) points out that stories of prophetic dreams, and dreams in which kings and holy men consciously interact with dream characters are as old as history. The North American Plains Indians' well documented vision quests were a special type of lucid dream in which a specific dream character took on the role of spirit guide for adolescent warriors. Many African cultures stress the importance of dream interaction with ancestors when making decisions that affect the family. Some Asian cultures also have complex systems of understanding dreaming, from the interpretation of mundane dreams, to the belief among some groups in India that dream consciousness is a higher state of reality than waking consciousness. Bynum concludes that the dynamics of lucid dreaming are "available to all individuals...a potentially extremely creative realm available not only for ideas of a scientific or an artistic nature, but also for the healing process as it relates to the psychophysical system of the body" (p. 217). In all of these cultures, the conscious interaction of the dreamer with the dream has the specific purpose of confronting and resolving the psychological conflicts that affect the person in waking life. In his research on lucid dreaming, Tart (1979) realized the benefit that could be gained from dealing directly with potential conflicts on the unconscious level as they arise. Being able to deal with them in the context of a lucid dream can help resolve them before they become overt problems in everyday life. REM sleep physiology Psychodynamic theorists are not the only psychologists interested in REM sleep. Mental disorders such as depression also exhibit REM abnormalities. One of the most common symptoms of depression is sleep disturbance, usually taking the form of insomnia or early risings, along with frequent awakenings throughout the night. In some cases the sleep disturbance can be excessive sleep, with the victim spending 12 to 15 hours a day asleep. Sleep disturbance is also manifest in abnormalities of the sleep cycle itself: [Depression] may have to do with biological rhythms... Depressives also consistently show abnormalities in their progress through the stages of sleep. One such abnormality is shortened REM sleep latency - that is, in depressives the time between the onset of sleep and the onset of REM sleep, the stage of sleep in which dreams occur, is unusually short. (Alloy, Acocella & Bootzin, 1996, p. 256) A comparison between dreaming and mental disorders is inevitable, given the dissociative quality and wild emotional fluctuations that are experienced by normal adults in dreams. According to Hobson: Dreaming is a model of psychopathology, a normal functional psychosis whose understanding could provide a key to major psychiatric dysfunctions. Thus dreaming (and REM sleep) reveals the mechanisms of the hallucinations and delusions of the so-called functional psychoses (schizophrenia and the affective disorders) and the disorientation and memory loss of the organic dementias. Like the mental status of schizophrenia, that of dreaming is also associatively loose. Like the mental status of the organic brain disease states, that of dreaming is also impressively confabulatory. In its emotional spectrum, dreaming often mimics mania and a panic anxiety state. (Hobson, 1988, pp. 230) Aside from the mental disorders, REM sleep also plays a role in maintaining the health of the immune system. An adrenal hormone known as dehydroepiandrosterone, or DHEA, is crucial to the production of antibodies. DHEA levels gradually decrease with age, and a recent study has shown that the effects of a disorder known as immune senescence, an immune system disorder that afflicts some elderly people, can be reversed by exposure to DHEA (Weksler, 1993). The hormone DHEA seems to be produced in its greatest quantities during REM sleep (Howard, 1995). There are several theories about how the brain regulates the REM/non-REM cycle. An hypothesis proposed by Howard (1995) involves a relationship between melatonin and DHEA. Melatonin is a sleep inducing hormone released into the blood by the pineal gland. The level of melatonin in the blood follows a circadian rhythm, with its lowest levels in daylight hours, and its highest levels (up to 10 times greater) at night (Glanze, Anderson, Anderson, 1996). This increasing level of melatonin in the blood in the late evening hours has an effect on overall body temperature, and induces sleep through thermoregulatory mechanisms. "By lowering core body temperature, melatonin reduces arousal and increases sleep-propensity. Thus, in humans, one role of melatonin is to transduce the light-dark cycle and define a window-of-opportunity in which sleep-propensity is enhanced" (Dawson and Encel, 1993, p. 1). DHEA, on the other hand, decreases during the day, and gradually builds back up at night at the same time melatonin levels are beginning to drop off. DHEA is used by the body to enhance metabolism, and so high levels of DHEA are associated with high levels of neural and physical activity. It is also one of the principal adrenal steroids used by the immune system. According to Howard's hypothesis, During sleep, reduced nighttime levels of DHEA reciprocate with melatonin to produce either slow-wave sleep (SWS) or REM sleep. These low productions of DHEA are higher during REM sleep, and lower during SWS. As sleep occurs, melatonin is "used up," and DHEA increases. During consciousness, DHEA is used up; melatonin increases, but is not released until DHEA levels decline to a low level prior to sleep. (Howard, 1995, p. 1) This hypothesis is partly supported by the fact that DHEA levels in the body gradually decrease with aging at roughly the same rate as age-associated changes in the immune system and sleep cycle take place (Weksler, 1993). Nevertheless, a cause and effect for the correlation of these hormones with the sleep cycle has yet to be demonstrated. A more well-established mechanism, which has been revised several times since it was first proposed in the mid-1970's, is the Reciprocal-Interaction model. This model is based on the interaction of three neurotransmitters: norepinephrine, serotonin, and acetylcholine. Norepinephrine and serotonin are biogenic amines, and they regulate neural activity by inhibiting the firing of the neurons they contact. Neurons that release these amines are called aminergic. Acetylcholine is an excitatory neurotransmitter - it causes neurons to fire. Neurons that release acetylcholine are called cholinergic. The Reciprocal-Interaction model describes the regulatory mechanism as a function of the relationship between neurons that increase the level of cholinergens and those that inhibit their release. The idea is that as the brain approaches REM sleep, the cells that inhibit the cholinergic cells are slowing their firing rate. When this inhibitory process reaches a low enough level, the cholinergic cells become active and REM sleep is initiated. According to Hobson (1988): The Reciprocal-Interaction hypothesis proposes that the continuous competition between the excitatory reticular neurons and the inhibitory aminergic neurons is the basic physiological process underlying sleep-cycle alternation. According to this theory, REM sleep and dreaming occur only when the activity in the REM-off aminergic neuronal population has reached a level low enough to allow the REM-on reticular system to escape from its inhibitory control. (p. 184) The neural mechanisms that control the sleep cycle lie in the brainstem. The non-REM (NREM) sleep mechanisms in the basal forebrain interact with the medullary and midbrain reticular systems to produce the slow EMG waves in the cortex. When the inhibitory neurotransmitters reach a low enough level, the excitatory acetylcholine neurons in the pons take over, initiating the REM stage and sending periodic signals to the cortex. These signals stimulate the neural activity that we experience as dreaming. Hobson's theory is supported by a recent study (Maquet, et al. 1996) of cerebral blood flow during REM sleep. Their study showed a positive correlation between regional blood flow in these regions of the brainstem and REM sleep. The amygdaloid complexes, thalamus, and certain parts of the parietal lobe showed increased blood flow, possibly accounting for the deep emotional content of dreams. As stated by Maquet, et al., Given the role of the amygdaloid complexes in the acquisition of emotionally influenced memories, the pattern of activation in the amygdala and the cortical areas provides a biological basis for the processing of some types of memory during REM sleep. (p. 163) Also supported by the study is one aspect of the memory- consolidation model of dreaming, although only specific types of memory are implicated: In man, REM sleep particularly influences the processing of procedural-implicit memory and also of memories acquired under emotionally charged conditions. Thus, the activation of the amygdaloid complexes suggests that REM sleep contributes to some memory processing. Moreover, the rCBF [regional cerebral blood flow] distribution suggests a functional link between the amygdala, the hippocampal formations, and cortical areas during REM sleep. (p. 164) The reticular formation in the brain-stem, like many other areas of the brain, is not fully understood. It contains the nerves that control the various stages of sleep, and groups of nerve bundles that communicate with parts of the forebrain. The reticular formation contains the locus ceruleus, which seems to play a critical role in dreaming. Francis Crick (1994) explains, ...a small group of neurons, called the locus ceruleus, sends signals to various places, including the cortex. A single one of these nerve fibers can extend from the front of the cortex to the back, making millions of connections along the way. The exact function of the locus ceruleus is unknown. It becomes almost completely inactive in REM sleep. It is possible that its activity is needed when the cortex needs to put a memory into long-term storage. Its inactivity in REM may help to explain why we are unable to recall the majority of our dreams. (p. 89) Exactly how the brain creates, processes, and retrieves memory is a question that is still being investigated. Many of the mechanisms are understood, and correlations have been established between certain areas of the brain and memory processing. But the neurological details, particularly those involved in both REM sleep and waking memory, have yet to be completely worked out. Theoretical studies Since much of the same neurological geography involved in memory is also active during REM sleep, it is natural that many studies of REM sleep have focused on the interaction between memory and dreams. Various theoretical models have been proposed, each of which faces the task of accounting for dreams. Some models have done so better than others. Theoretical models must also account for the fact that some people dream quite differently than others. A promising area of research into this difference is the study of field independence in individuals. The capacity for field independence seems to depend on how well both cerebral hemispheres can work together ("bilateral hemispheric symmetry" is the term usually encountered to describe coordination between the left and right hemispheres.) Studies of dreaming and memory Dreams, like most ordinary events during waking, seem to be processed in short-term memory. What appears to be weak in sleep is the path that leads from short-term to long-term memory. Many people can remember several dreams each night, but memories of dreams are sparse. It is rare that all of the events in a dream can be recalled. Usually only tiny fragments make it to long- term memory, and even these can be forgotten quickly during the day if the dream is not rehearsed or written down immediately upon awakening. The act of waking up itself floods short term memory with stimuli: the beeping alarm clock, the plans for the day, what to wear, it's no wonder that our dreams are usually forgotten the moment our eyes open. Also, the key to storing items into and retrieving items from long term memory is the availability of cues and reminders. While awake, we can remember the things we did yesterday or last week because we operate in the same environments day after day. These environments are full of reminders of past activities. It's easy to remember what we had for dinner last night because the dishes we used to cook it are still in the sink. We remember that we went to the beach last weekend because there are still a few grains of sand on the car seat. But what reminds us of last night's dream? Its events might stir grudgingly to memory if the dream evoked powerful emotions, or if by some chance during the day an event occurs that reminds us of it. But usually there is no continuity between our dreams and waking life, and without the availability of cues between our outer waking world and our inner dreamworld, dream recall can be difficult. In a 1991 review on the topic of dream recall, Donald Goodenough noted both primacy and recency effects in dream recall. Healthy individuals have several dreams every night, but it is almost always the last dream of the night that is most easily remembered in the morning. Even subjects in a sleep lab, who are awakened at each REM stage and asked to report their dreams, have more difficulty remembering earlier dreams than later ones. The only exception is the first dream of the night. While the first dream is not remembered as easily as the last, it is recalled more readily than the middle dreams. As Goodenough states, "serial effects may play an important role in determining which dreams of the night are recalled" (p. 162). The fact that serial learning theory can be applied to dream recall leads to questions about other types of dream learning. It has been demonstrated that classical conditioning transfers readily to the sleeping state (Beh and Barratt, 1965), and it is clear that long term memories can be retrieved during dreams. But can long term memories be created in dreams? Pietro Badia (1990) pointed out that while procedural knowledge (which includes habituation, sensitization, and classical conditioning) can be acquired in sleep, the ability to acquire declarative knowledge (facts, numbers, events, etc.) is apparently limited. A correlation has been established between the presence of alpha-waves in sleep and the ability to learn in sleep, with the implication that alpha-waves, normally associated with the waking state, appear to be necessary for the formation of long-term memories. The alpha-wave correlation is not contradicted by the phenomenon of lucid dreaming, in which declarative memory seems to be active. REM sleep sometimes produces alpha-waves. A 1982 study by Olgivie, et. al., showed that "arousals from high alpha REM sleep yielded significantly higher lucidity" (p. 795) than was found in dreamers awakened from low-alpha REM sleep. So if lucid dreams are correlated with high-alpha REM sleep, then it is possible that the connection between long-term memory and short-term "dreaming" memory is in fact active during these intense and memorable dreams. Also of interest is that, according to Badia (1990), people who experience anterograde amnesia exhibit nearly identical learning characteristics as have been found to occur in sleep learning. Amnesiacs can acquire procedural memory through classical conditioning or habituation, but they cannot acquire new declarative memories. Badia states that the critical brain site is the hippocampus. Lesions or injuries to the hippocampus disrupt the formation of new memories. Badia hypothesizes that, The hippocampus, in the presence of an EEG sleep pattern, is not functional for the acquisition of declarative knowledge; that is, it serves its neural gating function of transferring short-term to long-term memory only in the presence of an EEG waking pattern...Others have also noted that the hippocampal contribution to information flow is highly state dependent. (p. 74) Evidence for a "gating function" in the hippocampus is supported by learning studies conducted on rats. A study by Wilson & McNaughton (1994) measured the firing patterns of specific cell groups in the hippocampus of rats while the rats learned spatial behavioral tasks. It was found that neurons that fired together while the animals were learning tended to fire together during slow-wave sleep after the learning had taken place. These firing patterns had not been present before the rats learned the task. This study lends support to the memory-consolidation theories of sleep and dreaming. Wilson & McNaughton indicate that the repetition in sleep of neural patterns learned while awake "may result from synaptic modification during waking experience. Information acquired during active behavior is thus re-expressed in hippocampal circuits during sleep" (p. 676). Neural network models The most recent and complete theoretical models of mind and memory involve the concept of neural networks. In neural network models, knowledge and memory are seen as being stored in the connections between the millions of neural synapses in the brain. Sensory input stimulates a pattern of nerve firings, and this pattern activates other nerve groups throughout the cortex. These nerve groups respond with firing patterns of their own. Memories and associations are formed when the stimulus firing pattern consistently invokes specific response patterns. The interconnected neural cells store associations between events (knowledge) in the connection strengths between neurons. Crick & Mitchison (1983) developed one of the more well- known neural net models. In their model the role of dreams is not to consolidate memories, but rather, to "unlearn" certain maladaptive patterns of thought that can develop in overloaded networks. In this model (as in most other neural net models), memories stored as connections between neural synapses can be triggered by other memories or thought events that resemble the firing patterns that formed those memories in the first place. According to neural network theory, "If all the cells involved in an event form mutual synapses, then when part of that event is encountered again these synapses can cause the regeneration of the activity in the entire subset" (Crick & Mitchison, 1983, p. 111). In other words, once a memory is formed, the brain's neural network can recreate that memory with only a fraction of the original stimulus, e.g. a cue or a reminder. Imagine a multitude of interconnected memories formed over a lifetime of learning. Triggering one memory can create an internal mental stimulus (neuronal firing pattern) that in turn triggers another memory. In the neural network model, a single neuron can be involved in the firing patterns of many different memories. As Crick & Mitchison state, The associations [memories] which are stored are not assigned specific locations for each item, as in a digital computer. Instead the information is: (1) Distributed: this implies that a particular piece of information is distributed over very many synapses. (2) Robust: this implies that the information will not be totally lost if a few synapses are added or removed. (3) Superimposed: this implies that any single synapse is involved in storing several distinct pieces of information. (p. 111) What happens when such a network becomes overloaded? In simplified computer simulations of this model, overloaded neural nets tend to get stuck in repetitive patterns. In a human mind these patterns would be reflective of modes of behavior resembling psychoses. In Crick & Mitchison's words, (1) The net may produce many far-fetched or bizarre associations ('fantasy'). (2) The net may tend to produce the same state, or one of a small set of states, whatever the input ('obsession'). (3) Certain kinds of nets, particularly those which feed back on themselves, may respond to inappropriate input signals which would normally evoke no response from the net ('hallucination'). (p. 112) These maladaptive modes are called "parasitic," and once established, they tend to dominate the network. How can a neural network protect itself from parasitic modes? According to the network model, the best way to clean up parasitic modes is to close off both input and output, and randomly stimulate the network. When this is done repeatedly, with no learning responses from the net, it has a dampening effect on parasitic modes, causing them to reorganize and resolve themselves into healthier response patterns. As we have already seen, this is what happens when we're dreaming. During REM sleep signals from the brainstem shut down both sensory input and responsive output (sleep paralysis). The aminergic (inhibitory) neurons in the brain switch off, leaving the cholinergic (excitatory) neurons free to fire in response to any and all signals. The reticular formation in the pons begins sending periodic signals to the cortex, and the mental activity that results from this internal stimuli is experienced as dreaming. So dreams, according Crick & Mitchison, are side-effects of the brain's house-keeping mechanism. What we perceive as visual imagery and affective experience may be nothing more than "reverse learning," in which an overloaded network purges itself of parasitic modes. They go so far as to advise that "attempting to remember one's dreams should perhaps not be encouraged, because such remembering may help to retain patterns of thought which are better forgotten" (p. 114). Neural network models represents an important conceptual bridge between the physiological and psychological aspects of the mind. However, Crick & Mitchison's model has some problems with respect to its explanation of dreaming. First of all, if dreams are a response to overloading, why do infants spend so much more time in REM sleep than adults? The ideas mentioned earlier - that REM sleep in infants help promote neural development - fit the data better. As we grow older and accumulate more memories and experience, Crick & Mitchison's model would suggest that we need more REM sleep, not less. Yet the reverse is what we see in real life. Second, according to this model, dreams must be random and disorganized to effectively eliminate parasitic modes. Yet those who report frequent lucid dreams, such as long time meditators (Gackenbach & Bosveld, 1989) are as mentally healthy (if not more so) as those who experience the normal, random dreams demanded by the model. Finally, dreams often have a positive emotional effect. Clear and vivid dreams can provide new perspective on troubling issues, allowing us to confront problems or experience sensations in ways not possible in waking life (Walsh & Vaughan, 1992). Strong, positive emotional dream experiences do not support the idea that dreaming is simply a mental "reverse learning" mechanism. Rosalind Cartwright (1990) offers a different neural network model that better accounts for dreams. She notes that dreams tend to progress through the night from thought-like content early on, to more bizarre, vivid experiences later in the night when REM is highest. Rather than being random, unrelated bits of image and emotion, she has found that dreams tend to repeat the same images and thematic content from the beginning of the night to the end. She also notes that our emotions serve as connections between memories, and states that emotional concerns in our waking life carry over into sleep. So, while the REM state itself may be activated by random pings from the brainstem, the specific memory networks that are stimulated are guided by emotions. She writes, Memories are encoded as unique bits but are organized into networks of related materials and are decoded according to our needs...Emotional connections are made via earlier memories. These are then lit up selectively when the brain is activated in REM sleep. Dream meaning relates to the status of our ongoing needs, and dreams function to assimilate new data and reorganize related memories. Once the network is stimulated in the first REM period, it is held in a state of readiness, and other parts of it are more likely to be stimulated the next time REM occurs when older elements of the same network are stimulated. (p. 186) As the night progresses, each REM stage activates deeper and deeper layers of the same network, contributing to the sense of significance and personal meaning that often results from long, vivid dreams. Globus (1989) developed a neural network model that is similar to Crick & Mitchison's, but which accepts the rich emotional nature of dreams as an important element, not just a side-effect. As we have already seen, neural networks are self- organizing, self-consistent systems. Input (internal or external sensory stimuli) perturbs the system. The system, through its response to the input (chains of neural firing patterns set off by the input), works toward settling back to a harmonious state. This resolution sometimes involves some kind of output (behavioral response). Globus envisions three layers: (1) the perceptual layer, governed by input from the five senses by "tuning" (what we pay attention to), and by memory. The output of the perceptual layer is (2) the cognitive layer, which responds to perceptual input to find its own self-consistent solution (i.e. incorporate a new idea, confirm an existing idea, throw out a bad idea, or respond with a behavior of some kind.) The output of the cognitive layer then moves to (3) the behavioral layer, "which has its own tuning and knowledge, self-organizes toward harmony, and emits behavioral instructions" (Globus, 1989, p.192). In Globus' model, each of the three layers operates under certain constraints. Input is obviously one constraint - sensory input from the environment determines what the network experiences and learns over time. Another is knowledge, the particular associations and connection strengths that the network has developed over its lifetime. The third constraint, and the most interesting with respect to dreaming, is intent, or "tuning signals" in network parlance. When a network is said to settle into a harmonious state, it is intent that defines what that harmonious state should be. In Globus' (1989) words, Intentional acts are mental acts, such as perceiving, believing, desiring, and planning. Intentional acts are directed upon objects by meanings. Meanings "prescribe" objects, specifying what is to be found in the world in perception, and states of the world desired, believed in, planned for, and so on...The intentional act may or may not be "successful," depending on whether or not the world actually satisfies the prescriptive meaning. So by "intentionality" I mean mental acts that prescribe objects. In dreaming as in waking we perceive, believe, and act, engaging in a variety of intentional operations. (p. 188) Going back to Crick & Mitchison's (1983) network model, we see that their perception that dreams occur randomly does not take into account either the tuning signals (intent) that constrain the network or the emotional links between the memories that are activated in dreams. Meaning forms the context within which intent operates, and "it is meanings that are remembered in dreaming, not the events that are the objects of the intentional acts" (Globus, 1989, p.190). And it is "salient meanings left over from the day" (p. 192), both cognitive and emotional, that tunes (guides) the direction and content of our dreams. Field Independence and Dreaming The perspective a person has in both waking daydreams and sleeping dreams seems to correlate with the type of imagery that is experienced. Foulkes & Kerr (1994) found that when a daydream is experienced from an outside perspective, such as when you see yourself in a scene as a third party might see you, then the imagery tends to be static, like a single photograph. When the perspective is seen as though you were looking through your own eyes, then the daydreamed imagery is likely to be lively and more imaginative, as though you were in an active scene that is moving. It was also noted that implausible or unrealistic imagery was most likely to appear from the own-eyes perspective. Similarly in nocturnal dreaming, movie-like imagery is most often experienced when one is experiencing the dream from the own-eyes perspective. This may have to do with the field/background perspective, in that while one is looking around, it is the dreamed situation that occupies the attention. When viewing yourself, on the other hand, it is you that commands your attention. It may be that focusing on yourself somehow inhibits the mental processing of extraneous scenery or images, whereas focusing outward from the self while experiencing the dream frees the mind to create scenes and situations with which the dreamer can interact. As Foulkes & Kerr (1994) state, "The suggestion is that the see-oneself perspective is incompatible with the sustained production of involuntary kinematic imagery" (p. 690). A series of studies by Gackenbach, Heilman, Boyt, & LaBerge (1985) found that subjects who experienced frequent lucid dreams were indeed "significantly more field independent than both non- lucid dreamers and infrequent lucid dreamers" (p. 17). They also found an interesting gender difference in their measures of field independence. They used two measures, the embedded figures test (EFT) and the rod and frame test (RFT). Both male and female lucid dreamers outperformed their non-lucid peers in the EFT. However, only male lucid dreamers performed significantly better in the RFT. In the RFT, all groups of females (frequent lucid, occasional lucid, and non-lucid dreamers) performed equally well. What is the difference between these two measures of field independence? The RFT is a pure spatial-orientation task. The subject stands in a darkened room looking at a frame in which a rod is suspended. The rod can be slowly spun clockwise or counter-clockwise by an experimenter. The subject's task is to align the rod so that it is parallel to the wall, which must be done by instructing the experimenter to move the rod one way or the other. In the EFT, the subject must locate simple geometric shapes hidden within a complex design. While there is a positive correlation between EFT scores and scores on the RFT, the nature of the tasks is quite different. In the EFT, the subject can name the simple figure (e.g. "square" or "triangle"), and keep that verbal identification in mind while searching for it in the complex figure. This "verbal mediation" appears to activate both the left (verbal) and right (visual/spatial) cerebral hemispheres, suggesting that bilateral hemispheric symmetry may be a factor in distinguishing female lucid dreamers from the other female subject groups. Garrick (1978) states that "the dominant hemisphere seems more closely associated with recall and relearning while recognition is associated with both hemispheres" (p. 632). Recognition seems to be the key to doing well on the EFT. With respect to verbal mediation, Garrick also notes that the "inferiority of field-dependent subjects in identifying verbal concepts" (p. 636) may have to do with the fact that some verbal tasks have "perceptual and imagery components," (p. 636) requiring the complementary function of both cerebral hemispheres. Intense dreams Intense dreams can be distinguished from normal dreams in the same way that frightening or ecstatic experiences in waking life are distinguished from mundane events. The former are memorable and laden with emotion, sought after or dreaded. The latter are barely noticed and immediately forgotten. Three types of intense dreams have been studied: nightmares, archetypal dreams, and lucid dreams. Nightmares can be vivid, are often terrifying, and can leave the dreamer feeling upset and exhausted after awakening. Archetypal dreams generally involve some kind of self-transcendent experience, whether through encounters with mythical figures, or by experiencing dream imagery that conveys a rich sense of personal meaning. A lucid dream is concretely defined as "a dream in which you are actively thinking about the fact that you are dreaming" (Gackenbach & Bosveld, 1989, p. 186). Nightmares In general, females report more nightmares than males. Levin (1994) reports that those who suffer chronic nightmares are generally better at dream recall than non-nightmare control subjects, and also experience higher levels of aggression in their dreams. He states that "nightmare sufferers appear to be more internally directed and more sensitive to their generalized dreaming states" (p. 127). Other researchers (Berquier & Ashton, 1992) have found that lifelong nightmare sufferers "scored significantly higher on the Eysenck Personality Questionnaire Neuroticism scale and on 8 MMPI clinical scales" (p. 246). Their study indicates that nightmares may result from global maladjustment rather than from any specific psychosis. Physiological confounds may exist as well. Bearden (1991) notes that chronic nightmares, post-traumatic stress nightmares, and drug-induced nightmares all "appear to be related to a hyperreactivity of the nervous system which may have the effect of increasing REM pressure" (p. 139). Archetypal dreams Archetypal dreams are characterized by emotions of awe and wonder. These dreams have two general forms. The most obvious is one in which the dreamer encounters images or dream characters that inspire feelings usually associated with deep religious experience. The figures and characters experienced have a mythological quality to them, which the dreamer may associate with deep personal meaning. Archetypal dreams can also take the form of geometric patterns and images. Geometric shapes such as checkerboard patterns, crystalline latices, or moving lines appear as a bright, shifting image before the eyes. The dominant feature of these dreams seems not to be the geometric patterns, but rather, the quality of the light in which they are experienced. George Gillespie described the kind of archetypal dreams in which "stable, intense lights" (p. 488) make their appearance. He characterized these phenomena as "persisting, unchanging bright lights...that do not behave like ordinary dream images" (p. 488). For one thing, they remain visible in the same location in the visual field no matter which direction the (dreamed) head is turned. They don't seem to be a part of the dream imagery. And unlike other dream forms and images, they do not appear to be representational. The lights appear in several forms. One is simply a vaguely defined area of light that blocks out other dream imagery that may be present. Another is peripheral light that appears (and stays) just outside the field of vision. Various sized disks of light, with clearly defined borders, might also appear, and sometimes a "sun-like concentration of light" (p. 489) is visible. Gillespie states that such concentrations are qualitatively different than dreamed images of the sun itself, and that at times several of these sun-like lights appear together. The most interesting light phenomenon Gillespie calls "fullness of light," (p. 489) in which the entire visual field becomes filled with bright, intense light. This light is accompanied by feelings of mystery or even sacredness that is not present in the other light phenomenon. He states that "during the [fullness of] light I feel intense joy and devotion. I believe at the time that God is present in the light" (p. 489). These dream images differ from normal dream images in that they do not appear to represent anything, and seem to be unrelated to other dream images. The lights "do not act in coherence with other dream images to make a coordinated whole" (p. 491). While the lights maintain their position within the visual field, they can be scanned. One can look toward them or away from them. They can also have the effect of eidetic imagery, that is, imagery that remains visible for a short while after one has awakened. When this occurs, the lights still maintain their position within the field of vision, and unlike sunspots or afterimages, they can be still be scanned. Some "lights" dreams do not involve normal dream imagery at all, but rather, patterns of geometric shapes. These shapes can appear against a background of light, within a diffuse glow of light, or themselves comprised of filaments of light. These geometric patterns can be seen as either flat and two- dimensional, or as three-dimensional images. Spadafora and Hunt (1990) found a correlation between the occurrence of these kinds of dreams and high scores on tests which measure a subject's tendency to mentally animate simple line drawings. This capacity is "a key process in the capacity for metaphor" (Hunt, 1989, p. 633). If dreams involving light and geometric patterns occur in individuals with high capacity for metaphor, then perhaps the dreams themselves provide insight on the transformation of abstract form into personal meaning. Background to Lucid Dreaming In 1913 Dutch psychotherapist Frederik van Eeden published a paper simply titled A Study of Dreams. Van Eeden described nine different kinds of dreams, and is credited with coining the term "lucid dream." The paper described and classified many of the phenomenon that dream researchers are now familiar with, such as flying, conversing with dream characters, or believing that you have awakened only to discover that you're still dreaming. Unfortunately, the paper suffers from many fallacies reflective of the thought and culture of the Victorian era. One type of dream is described as being influenced by the devil, for example, and another type as being the result of illness or poor digestion. Yet another type - the false awakening dreams - were considered by Van Eeden to be demon pranks. Despite these questionable classifications, the paper is nonetheless considered to be a classic. Where dreams come from has produced a great deal of speculation among philosophers and scientists alike. Freud's ideas that dreams originate in the unconscious have been among the most enduring. Freud believed the mind to be comprised of three layers - the conscious, the preconscious, and the unconscious. The unconscious, inaccessible to waking consciousness, expresses its repressed desires through nocturnal dreams. In 1982 Rosalind Cartwright devised an experiment to determine whether dreams really came from the unconscious, or were instead drawn from material in the preconscious. According to Freud, the preconscious is that part of the ego in which "internal events such as passages of ideas and thought-processes can become conscious" (Freud, p.35). Freud characterized the preconscious as containing residues of speech (p. 86) and mnemonic residues of visual and auditory perceptions (p. 35). The preconscious, then, is thought by some to the source of the self-talk and fantasy that underlies most of our awareness. As Cartwright put it, "It is generally conceded that some preconscious processing of feelings and thoughts is continuous throughout waking and is difficult to suppress whenever little or no concentration is required" (Cartwright, 1982, p. 433). Cartwright's hypothesis was that "the thematic content of relaxed wakefulness, when mental life is under moderately low demand conditions and conscious control, is draw from the same source as dream content and that both differ from thoughts reported under high demand or subjective control" (p. 434). In other words, when we aren't thinking about anything in particular, and our minds are allowed to wander, the seemingly random stream of thoughts we experience arise from the same source (the preconscious) as dreams. To test this hypothesis, Cartwright had her subjects sit quietly for an hour before going to sleep. At random intervals the subjects were asked what was presently going through their minds. During the night, the subjects were awakened after each period of REM, and their dreams recorded. The following day, the subjects were interviewed about life events that were important to them at the moment. It was found that what was going on in a subjects life had little relation to that subject's dreams. However, the themes that arose in their minds during the period of relaxed wakefulness were correlated very highly with the content of their dreams. This study stemmed from an earlier work that explored the possibility that people could choose what they dream about. Her 1974 study was one of the first to clearly demonstrate that dream content can be influenced by conscious desire. Cartwright gave each of 17 subjects a word that described some aspect of that subject's personality that the person wanted to change. She also selected control words for each subject. She had her subjects repeat to themselves the word they were given as they drifted off to sleep. The subjects were then awakened during the night after each REM period, and any dreams they described were recorded. Finally, Cartwright provided independent judges with a list containing the target words and the control words, and asked them to match the words to the dream reports. The goal was to see if the subjects dreamed about the target word and not the control word. It was found that this in fact happened for a significant number of subjects. Rosalind Cartwright's work provided the scientific foundation that was needed for dream research to be taken seriously. The work done by Saint-Denys, Van Eeden, and even the dream accounts by Freud only amounted to anecdotal evidence that dreams were anything more than random events. Lucid Dreaming Research on gaining control of dreams and dream content soon became focused on lucid dreaming. One of the first papers to describe a technique for reliably inducing lucid dreams was written by Stephen LaBerge in 1980. It was a case study, with himself as the subject, describing a lucid dream induction technique that he had developed. He based the initial part of his study on comments made by Patricia Garfield (1975), who stated that autosuggestion could be used to induce 4 or 5 lucid dreams per month. LaBerge spent the first 16 months of the three year study practicing autosuggestion. During this period he isolated two factors that seemed to be associated with the occurrence of lucid dreams. The first was motivation, and the second, "the pre-sleep intention to remember to be lucid during the next dream" (LaBerge, 1980, p.1041). With the clarification of these factors, LaBerge was able to develop a technique that he called the "mnemonic induction of lucid dreams," or "MILD" (p. 1041). The MILD technique has five steps. The first is to learn to awaken in the early morning hours when the amount of REM sleep is at its highest. Once awake, you should engage in 15 minutes of activity, such as reading, working a crossword puzzle, or anything that demands alert wakefulness. LaBerge based this step on reports that lucid dreams often followed activities in the middle of the night, and found it to be effective in his research as well. Some researchers (Sparrow, 1976; Gackenbach & Bosveld, 1989) suggest that an early morning meditation session is particularly helpful. The third step is to return to bed, and while returning to sleep repeat to yourself, "Next time I'm dreaming I want to remember that I am dreaming" (p. 1041). Then (fourth), visualize your sleeping body, and the rapid eye movements that indicate dreaming. While visualizing this, see yourself in a familiar dream, either one you may have awakened from earlier or one that you remember from a previous night. Finally, continue steps three and four as you fall asleep. This appears to fix the intent to have a lucid dream firmly in the mind. Using the MILD technique, LaBerge was able to increase the frequency of his lucid dreams from five or six per month (using autosuggestion) to over 21 per month, with as many as four in a single night. While few deny that lucid dreaming is a possibility, the subjective nature of the experience makes it difficult to verify that lucid dreams do in fact occur. Dream reports from reliable people claiming to have had the experience do not necessarily constitute proof that the event took place as described. No matter how reliable the source, dream reports still amount to little more than anecdotal evidence. They provide nothing that a disinterested observer can measure or replicate. In 1981 a study published by LaBerge, Nagel, Dement, & Zarcone provided definitive proof that subjects could not only take control of their dreams, but through subtle body signals could communicate back to the waking world from within their dreams. In this study, subjects experiencing a lucid dream were able to send coded messages from within the dream back to researchers in the lab in which the subjects slept. The subjects were monitored using an EOG to monitor eye movements, an EMG for muscular movements (attached to the wrists), and an EEG to measure brain waves. When the subjects entered into a lucid dream, they performed two prearranged signaling tasks. The first was to look upward with the dream eyes, a movement which has been found to affect eye movement in a subject's sleeping body. Looking upward in the dream, or moving the eyes back and forth in a deliberate sequence, create corresponding movements in the sleeper's eyes, and these movements can be seen on the EOG. Following the pre- arranged eye movements, the subject in the dream squeezed the left and right fists in a morse-code sequence (left fist for dot, right fist for dash) to spell out the subject's initials. These movements also produce visible patterns, this time on the wrist- attached EMG. In 24 out of 30 reports of dream signaling by the subjects, the results showed up clearly in the polygraphs. Shortly after the EEG showed a REM sleep pattern, the EOG recorded the prearranged eye movement. Shortly after the eye movement, the left and right wrist EMG graphs showed muscle activity in a sequence that corresponded to the subject's initials. Understanding Intense Dreams Often individuals who experience one type of intense dream also experience one of the other types as well, something that has proven problematic for researchers attempting to correlate each of these types of dreams (and dreamers) with particular physiological or personality characteristics. If a study attempts to isolate the characteristics of people who suffer from nightmares, for example, their subject population is likely to include people who also experience lucid dreams. This overlap makes it difficult to isolate characteristics that correspond to a specific type of intense dream. In 1990 Spadafora and Hunt published the results of a pair of studies that provided insight on the characteristics of individuals experiencing each of the three types of intense dreams. In the first study, their initial hypothesis was simply that intense dreamers would all score high on measures of creativity and imagination compared to a control group. This study showed that the three types of intense dreams were indeed correlated with each other. The study confirmed previous research, but it also allowed them to identify subjects who generally experience only one type of intense dream. In the follow-up study, Spadafora and Hunt selected subjects from the first study who experienced one of these types of dreams and not any of the other types. They added a fourth group of individuals (control gp) who were high in dream recall, but who did not experience intense dreams. Their goal in this study was to isolate each type of dreamer and try to identify distinguishing characteristics. The characteristics they measured were those that have been reported in the literature as being correlated with individuals who experience intense dreams. Subjects in each group were tested for imagination and creativity using a variety of scales, including imaginative absorption (tendency to become absorbed in an activity such that the surroundings are forgotten), creativity, a physiognomic cues test, and Hartmann's thin boundaries questionnaire. Another characteristic was physical balance, which is used to measure the functioning of the vestibular nerves in the brain stem, which appear to have an effect on REM sleep and lucidity. According to Gackenbach and Bosveld (1989), "Research indicates that these nerves contribute to the production of REM...In fact, these nerves are particularly active in the phase of REM that is associated with lucid dreaming" (p. 171). Also measured were the subjects' proclivity toward mysticism (a measure of the tendency to have peak mystical experiences), spatial-analytic abilities, stress, ordinary dream bizarreness, and archetypicality. The last two characteristics were identified by scoring entries in dream journals kept by the participants. Lucid and archetypal dreamers differed from the nightmare group in that they scored higher on the combined imagination- creativity scale, spacial-analytical abilities, and physical balance. One of the more interesting findings of this study was that when the imaginativeness scale was taken by itself, people who scored high in imaginativeness were "significantly high on ordinary dream bizarreness, archetypal dream estimates, and dream diary archetypicality" (p. 637). Contrary to earlier studies, however, they found no relation between imaginative absorption and either lucidity or nightmares. As was noted earlier, archetypal dreams have a metaphoric component that may not be necessary for lucid dream control and awareness. The cognitive reshuffling of images into new systems of meaning demands imagination. There is no certainty about whether mythic dream characters and symbols are themselves meaningful, or simply become meaningful as a result of the dreamer's imaginative combining of symbols and meaning. While lucid dreamers do score high on the composite creativity-imaginativeness score, other characteristics may play a more important role in dream lucidity. For example, Prescott & Pettigrew (1995) found a high correlation between lucid dreaming and control over situations in waking life. Since general self- awareness is necessary for self-control, people who maintain a high level of awareness while awake may be more likely to maintain that awareness while dreaming. So dream awareness (lucidity) may be more directly related to self-control than to creativity or imaginativeness. Spadafora and Hunt also found that nightmares are negatively correlated with spatial-analytical ability and proclivity to mystical experience. Subjects experiencing nightmares also scored significantly lower than the archetypal dreamers in the physical balance measures. Although physical balance correlated well with dream bizarreness, Spadafora and Hunt suggested that "subjects with good balance can better tolerate the imaginal reorganizations of visual-spatial structures" (p. 637). Lack of such tolerance may be the reason why this type of dream intensity is experienced in the form of nightmares, rather than the emotionally positive lucid or archetypal dreams. The picture that emerges is that intense dreams are experienced "positively or negatively, depending on variations in these cognitive dimensions" (p. 642). The low scores for mysticism and imaginativeness in the nightmare group "suggest a defensive inhibition of imaginativeness, where 'letting go' ends in panic rather than bliss" (p. 639). What remains unclear is to what extent those who experience nightmares can change their experience. Some who report nightmares also report attaining some lucidity as well, usually in the context of desperate efforts to wake up from a situation they have realized is only a dream. In 1991, Krakow, Kellner, Pathak, & Lambert used some of the dream-control techniques now available to psychologists to attempt to treat people suffering from nightmares. The treatment consisted of three steps: (1) write down a detailed description of a frequent nightmare; (2) change the events or characters in the nightmare to something more positive, or less threatening; and (3) rehearse these images as often as possible. The treatment was quite successful. The frequency of nightmares in the subjects decreased, and overall sleep quality improved. Conclusions Neurophysiology has taught us a lot about the brain mechanisms that initiate and control REM sleep. The cyclic interaction of aminergic and cholinergic neurons, regulated by signals from the pons in the brainstem, provide the engine that generates our dreams. Hormone production cycles also appear to work around the sleep cycle. Rising melatonin levels help induce sleep, and REM sleep appears to help replenish the body's DHEA, a hormone crucial to the immune system. When melatonin levels are highest, DHEA levels are at their lowest, and vice versa. When something goes wrong with our brain chemistry, such as when serotonin deficiencies associated with major depression are experienced, the sleep and dream cycle is disrupted. EEG studies demonstrate that alpha-waves appear to be necessary for long term memory to function efficiently. Long term memory seems to be active in high-alpha REM sleep, as demonstrated by the increased likelihood of lucidity during high- alpha REM. But the hypothesis that long-term memory is diminished or impeded in low-alpha REM is somewhat disconfirmed by the fact that mundane dreams (low-alpha REM) can also access long-term memory. Ordinary dreams clearly have access to long- term memory since it is the content of long-term memory (past events, current personal issues) that is dreamed about. Despite the characteristic dream distortions of places or events, long term memory seems to be just as active in low-alpha REM sleep as in high-alpha REM. Clearly there are memory mechanisms involved in sleep that have yet to be fully understood. Neural network models of the mind may provide insight on how memory and dreams interact. The work of Francis Crick and of Gordon Globus address neural networks from two different perspectives. Crick finds no meaning whatsoever in dreams, regarding them as the side-effects of mental house-cleaning. Globus, on the other hand, views dreams as a crucial key to understanding the mind. Globus' connectionism model dovetails well with both the physiological mechanisms of the brain and with the holistic concepts of psychologists such as Daniel Goleman, Ken Wilber, and others. The three parts of the connectionist model - environment (input), intent (tuning signals), and knowledge (connection strengths) - are supported by Cartwright, LaBerge, and others' findings that dream content can be controlled by knowledge and intent. Globus' theory is also supported by the meditation studies done by Daniel Goleman (1988) that indicate that the preconscious source of dreams (Cartwright, 1982) can not only be accessed but actively controlled. Most memory theories envision a clear path from short-term memory to long-term memory, with information traveling from an input source (e.g. the eyes), through this or that brain location to some final destination in the cortex where it can be retrieved later. The implication of neural network theory is that the mechanisms of memory may not follow such simple, step-by-step procedures. Rather, the brain operates as a holistic unit to process information and establish new memories. Input rarely arrives from a single source. When we experience an event, we usually see it, hear it, feel it, taste it, or smell it all at once. The event might be pleasant or painful. It might come in accordance with our expectations, or it might come as a complete surprise. The input from several different senses, often also accompanied by emotional sensations, combine to create a single event. When the event is dreamed about, the emotional content and bizarre re-associations surrounding the event in the dream seem better explained by neural network theory than by discussions about paths between short and long term memory. When a dream activates a specific layer of the neural network in the absence of our normal, waking inhibitions, associations with any and all other parts of the network are possible. The correlation between field-independence and the kinds of dreams people experience leads to some interesting questions. Can training in balance lead to healthier dreams? Can one improve the quality of one's dreams by practicing spatial relationship puzzles? It would be interesting to study the effect that gymnastics or art classes might have on dreams. Recent therapies for nightmare sufferers, and the techniques developed for inducing lucid dreaming, demonstrate that learning to think or perceive differently while awake can carry over into sleep. Since relatively few therapies have arisen from the study of dreaming, it might be promising to investigate the connection between waking training and dream experience. Modern dream research is a very young field. Throughout human history dreams have been connected to mystical phenomenon, and often accorded spiritual significance. Nietzsche (1878) identified the dream as the ultimate source of mysticism: The man of the ages of barbarous primordial culture believed that in the dream he was getting to know a second real world: here is the origin of all metaphysics. Without the dream, one would have had no occasion to divide the world into two. The dissection into soul and body is also connected with the oldest idea of the dream, likewise the postulation of a life of the soul, thus the origin of all belief in spirits and probably also of the belief in gods. 'The dead live on, for they appear to the living in dreams': that was the conclusion one formerly drew, throughout many millennia. (p. 54) Of particular interest is that many of those "primordial cultures" do hold concepts that are similar to the ideas of modern psychology. Freud described three layers of human consciousness: the conscious, the preconscious, and the unconscious. Anthropologist Carlos Castenada (1984, 1993) described a similar understanding of human awareness that he encountered in the Yaqui Indian shamans of northern Mexico. They described these layers as the known, the unknown, and the unknowable, and provided Castenada with a rich tradition of shamanistic practice aimed at gaining control of these layers. Two of their key practices were to learn to control their dreams, and to learn to master intent. While Castenada's work is regarded as questionable by some researchers, the dream-control techniques he outlined have been widely quoted in the dream literature by LaBerge, Gackenbach, Hunt, and others. And as we have seen, intent is a critical part of Globus' neural network model. Dream research is changing our theoretical understanding of consciousness, but it is unclear whether useful therapies can be developed from it. Nightmare sufferers have benefitted from dream-control techniques, but there has been little application to problems not specifically related to dream or sleep disruption. Part of the problem may be that dreaming is an entirely subjective experience. We do have polygraph records obtained from highly proficient lucid dreamers (whose lucid dreaming ability is not characteristic of the general population). But in general there is little evidence that dream techniques can be useful to those who rarely, if ever, experience lucidity. There is no doubt that positive dream experiences can have a beneficial emotional impact on waking life. Our understanding of the physiological mechanisms of dreaming, coupled with the an ever-increasing understanding of the psychological aspects of sleeping consciousness, may yet prove therapeutic in many ways. An important key to waking psychological health might very well lie hidden in the largely unexplored territory of sleep and dreams. ------------------------------------------------------------------ ------------------------------------------------------------------ References Alloy, Lauren B., Acocella, Joan, & Bootzin, Richard R. (1996). Abnormal Psychology. New York: McGraw-Hill. Badia, Petro (1990). Memories in Sleep: Old and New. In Bootzin, Richard R., Kihlstrom, John F., & Schacter, Daniel L., (eds.) Sleep and Cognition. Washington D.C.: American Psychological Association. Bearden, Carrie (1991). The Nightmare: Biological and Psychological Origins. Dreaming Journal of the Association for the Study of Dreams, 4(2), 139-152. Beh, H. C., & Barratt, P. E. H. (1965). Discrimination and conditioning during sleep as indicated by the electroencephalogram. Science, 147, 1470-1471. Berquier, Anne, & Ashton, Roderick (1992). Characteristics of the Frequent Nightmare Sufferer. Journal of Abnormal Psychology, 101(2), 246-250. Bloom, Floyd E., & Lazerson, Arlyne (1988). Brain, Mind, and Behavior. New York: W. H. Freeman and Company. Bynum, Edward Bruce (1993). Families and the Interpretation of Dreams: Awakening the Intimate Web. New York: Haworth. Cartwright, Rosalind (1974). The Influence of a Conscious Wish on Dreams. Journal of Abnormal Psychology, 83, 387-393. Cartwright, Rosalind (1990). A Network Model of Dreams. In Bootzin, Richard R., Kidhlstrom, John F., & Schacter, Daniel L., (eds.) Sleep and Cognition. Washington D.C.: American Psychological Association. Castenada, C. (1984). The Fire From Within. New York: Simon & Schuster, Inc. Castenada, C. (1993). The Art of Dreaming. New York: Harper Collins. Crick, Francis, & Mitchison, Graeme (1983). The Function of Dream Sleep. Nature, 304, p. 111-114. Crick, Francis (1994). The Astonishing Hypothesis. New York: Simon & Schuster. Dawson, D., & Encel, N. (1993). Melatonin and Sleep in Humans. Journal of Pineal Research, 15(1), 1-12. Folkes, David, & Kerr, Nancy H. (1994). Point of View in Nocturnal Dreaming. Perceptual and Motor Skills, 78, 690. Freud, Sigmund (1949). An Outline of Psycho-Analysis. New York: W. W. Norton & Company, Inc. Gackenbach, J., Heilman, N., Boyt, S., & LaBerge, S. (1985). The Relationship between Field Independence and Lucid Dreaming Ability. Journal of Mental Imagery, 9(1), 9-20. Gackenbach, J., & Bosveld, J. (1989). Control Your Dreams. New York: Harper & Row, Publishers. Garfield, P. (1975). Psychological concomitants of the lucid dream state. Sleep Research, 8, 153. Garrick, C. (1978). Field Dependence and Hemispheric Specialization. Perceptual and Motor Skills, 47, 631-639. Glanze, W., Anderson, K., & Anderson, L. (1996). Mosby Medical Encyclopedia (4th ed.). New York: Penguin Books USA, Inc. Gillespie, G. (1989). Lights and lattices and where they are seen. Perceptual and Motor Skills, 68, 487-504. Globus, Gordon G. (1989). Connectionism and The Dreaming Mind. The Journal of Mind and Behavior, 10, p. 179-195. Goleman, Daniel (1988). The Meditative Mind. New York: G. P. Putnam's Sons. Goodenough, Donald R. (1991). Dream Recall: History and Current Status of the Field. In Ellman, Steven J., & Antrobus, John S., The Mind in Sleep. New York: John Wiley & Sons, Inc. Hobson, A., & McCarley, R. (1977). The Brain as a Dream State Generator: An Activation-Synthesis Hypothesis of the Dream Process. American Journal of Psychiatry, 134, 1335-1348. Hobson, J. Allan (1988). The Dreaming Brain. New York: Basic Books. Hobson, J. Allan (1992). A New Model of Brain-Mind State: Activation Level, Input Source, and Mode of Processing (AIM). In Antrobus, John S., & Bertini, Mario (eds.), The Neuropsychology of Sleep and Dreaming. New Jersey: Lawrence ErlBaum Associates, Publishers. Howard, J. M. (1995). Online document. Available: http://gator.naples.net/~nfn03605/dheaslee.htm. Kihlstrom, J. F., Barnhardt, T. M., & Tataryn, D. J. (1991). Implicit Perception. In R. F. Bornstein & T. S. Pittman (Eds.), Perception without awareness. New York: Guilford Press. Krakow, B., Kellner, R., Pathak, D., & Lambert, L. (1991). Imagery Rehearsal Treatment for Chronic Nightmares. Behaviour Research and Therapy, 33(7), 837-843. LaBerge, Stephen P. (1980). Lucid Dreaming as a Learnable Skill: A Case Study. Perceptual and Motor Skills, 51, 1039-1042. LaBerge, S., & Rheingold, H. (1990). Exploring the World of Lucid Dreams. New York: Ballantine Books. LaBerge, Stephen (1992). Physiological Studies of Lucid Dreaming. In Antrobus, John S., & Bertini, Mario (eds.), The Neuropsychology of Sleep and Dreaming. New Jersey: Lawrence ErlBaum Associates, Publishers. Levin, Ross (1994). Sleep and Dreaming Characteristics of Frequent Nightmare Subjects in a University Population. Dreaming Journal of the Association for the Study of Dreams, 4(2), 127- 137. Maquet, P., Peters, J., Aerts, J., Delfiore, G., Degueldre, C., Luxen, A., & Franck, G. (1996). Functional Neuroanatomy of Human Rapid-eye-movement Sleep and Dreaming. Nature, 383, 163- 166. Nietzsche, F. (1878). Human, All Too Human. Excerpt quoted in A Nietzsche Reader. London: Penguin Books. Olgivie, Robert D., Hunt, Harry T., Tyson, Paul D., Lucescu, Melodie L., & Jeakins, Daniel B. (1982). Lucid Dreaming and Alpha Activity: A Preliminary Report. Perceptual and Motor Skills, 55, 795-808. Prescott, James A., & Pettigrew, C. Gary (1995). Lucid Dreaming and Control in Waking Life. Perceptual and Motor Skills, 81, 658. Rados, R., & Cartwright, R. (1982). Where Do Dreams Come From? A Comparison of Presleep and REM Sleep Thematic Content. Journal of Abnormal Psychology, 91, 433-436. Rosenzweig, Mark R., Bennett, Edward L., & Diamond, Marian C. (1972). Brain Changes in Response to Experience. Scientific American, 226, 22-29. Spadafora, A., & Hunt, H. (1990). The Multiplicity of Dreams: Cognitive-Affective Correlates of Lucid, Archetypal, and Nightmare Dreaming. Perceptual and Motor Skills, 71, 627-644. Tart, Charles T. (1979). From Spontaneous Event to Lucidity: A Review of Attempts to Consciously Control Nocturnal Dreaming. In Wolman, Benjamin B. (Ed.), Handbook of Dreams. New York: Van Nostrand Reinhold Company. Tholey, Paul (1983). Techniques for Inducing and Manipulating Lucid Dreams. Perceptual and Motor Skills, 57, 79- 90. Van Eeden, F. (1913). A Study of Dreams. Proceedings of the Society for Psychical Research, Vol. 26. Online document. Available: http://www.lucidity.com/vanEeden.html. Walsh, Roger N., & Vaughan, Frances (1992). Lucid Dreaming: Some Transpersonal Implications. The Journal of Transpersonal Psychology, 24(2), 193-200. Weksler, M.E. (1993). Immune Senescence and Adrenal Steroids: Immune Dysregulation and the action of dehydroepiandrosterone (DHEA) in old animals. European Journal of Clinical Pharmacology, 45 Suppl 1:S21-3; discussion S43-4. Wilson, Matthew A., & McNaughton, Bruce L. (1994). Reactivation of Hippocampal Ensemble Memories During Sleep. Science, 265, 676-682. ************************************************************** copyright © 1998 Dan Lee, all rights reserved.